Researchers could freelance and sell to such "brokering" companies, never worrying about products.
I doubt they can freelance and survive on a once-a-month part-time gabfest at Myhrvold's office. That only works because the tax payer and general population is paying most of the cost of generating these ideas in the first place: universities and university salaries.
Having somebody floating around whose only real motivation is ferreting out such scum and getting them to pay for the hard work that they're trying to sponge off of is a good thing.
Have you ever written any substantial piece of code? Chances are that you are infringing dozens of patents. Are you "scum" or a "sponger" because of it? I don't think so, since I doubt you even know of the existence of the patents you are infringing.
Many ideas that are being patented are so obvious that many people have them independently. The one who happens to be first to the patent office wins.
The real scum are the people who patent things that they know full well (or should know) are part of the public domain: ideas that others have talked about, ideas that have been discussed, ideas that are in textbooks. They steal from the public ideas and property to the tune of billions of dollars.
What I find interesting is how far we have come from a real free market. Ask any student of economics... what we have now in America and Europe is not a free market, rather there is government intervention propping up the strong and upsetting the "natural" balance that we would find with free market forces.
Governments need to intervene in markets in order to keep them free: it's only government intervention that can ensure that contracts are enforceable, that companies don't collude, and that people get the information they need.
The problem is that with the power to intervene for the purpose of keeping markets free, governments are also tempted to intervene for the purpose of lining the pockets of their private supporters.
Patents used to require hard work, experimentation, and demonstrating that something actually works. These days, Nathan invites some buddies for a "gabfest" and writes up the results as dozens of patents. Nobody cares whether it actually works, it just needs to contain a high enough proportion of patents that can be used to pressure companies to pay in order to pay for itself and make a handsome profit. And for a few thousand dollars a pop paid to his buddies for ideas that people can generate in a few minutes, Nathan's company gets to control millions of dollars of investments.
Unfortunately, many of the usual suspects are a member of this little club. Fortunately, if this company can do it, lots of other companies can do it, too. And if one of their client is being sued by another patents-only company, it doesn't matter how big their portfolio is: cross-licensing won't get them out of their legal troubles.
It seems Opera is growing, and they are doing it by aggressively promoting their products, even goes as far as teaming up with IBM's ViaVoice to allow users execute commands by talking to their computers. These are licensed-features that free browsers will find it hard to justify paying for.
The reason open source browsers don't have voice-guided browsing is because it's a useless gimmick. If there was any kind of demand for it, there are multiple open source speech recognition systems that it could be based on; nothing to license.
Are there any other "licensed-features" you can think of?
If you want to climb on a horse over the implementation of the interface, remember that javascript works on more than IE alone.
I'm not "climbing on horses", I'm having problems with Gmail on many of the devices and browsers that I use.
Most 'simpler' handhelds do Pop3 nowadays so they are/will be covered.
You want to look at a 500M mailbox with a handheld over POP3? I think you are a bit naive.
Furthermore, if they would like to get on the Geeks good side they should make a XUL version of their interface. That would create the killer-app for XUL in one go.
Yes, it would be the "geeky" thing to do. It just wouldn't be the sensible thing to do.
Apple's Spotlight isn't technology, it's a software product. The technology itself is various indexing, search, and user interface techniques that have been around for many years.
Gmail very un-Googlish
on
The Webmail Wars
·
· Score: 4, Insightful
The biggest problem with the Gmail interface is that it uses so much JavaScript; it doesn't run at all on many handheld browsers. I think at least they need a simplified XHTML interface.
The POP feature also makes no sense to me; it basically begs you to download messages from their servers when their stated goal is to collect lots of mail. If they offered IMAP access instead, people could keep their messages on the server. They could even use IMAP for placing subtle adds (e.g., message "1" is always some kind of simple ad, but unlike spam or hotmail, there would always only be one advertising message).
An area where Gmail could really do something better is passwords: they really should offer one-time passwords for travelers. Right now, when traveling, there is a high chance with web-based mail that your password gets compromised.
In any case, for fairly little money, you can get large mailboxes with IMAP interfaces from other companies, and you get a lot more control over them than with Gmail. Currently, Gmail's "free" isn't good enough for me to save the money I get with a commercial provider.
You missed the entire point that I was trying to make. Yes, "most people" will like PalmOS better because it has a better UI. But the system software that underlies that UI is horrid: the limitations and differences from commonly used operating systems mean that porting and developing for it is a lot of work and usually requires starting from scratch.
Palm has been able to get by on their better UI for years, but the poor job they have been doing on the underlying system software is catching up with them. And with 640x480 PDAs becoming more common, PPC's poor UI matters less and less anyway.
I don't see GPL violations becoming a big problem. First of all, the loss to the open source community from abuses of the GPL is mainly the lack of contributions. That is partially balanced out by the fact that the company in question is at least not using some proprietary solution. Second, the problem is self-limiting: if a company gets to be large enough, someone is going to notice the GPL violation and the company will likely settle the resulting legal action quickly.
For BSD-licensed software (are there any BSD VoIP solutions?), companies are even welcome to use the software and make it proprietary. Proponents of BSD probably believe that it's better to have companies use open-source derived software even if they don't give back than to lose more companies to proprietary vendors. (Personally, I think it needs to be looked at on a case-by-case basis.)
This one Palm did to themselves. Palm had half a dozen years to turn PalmOS into a modern, reliable 32bit operating system, instead they are still shipping handhelds that emulate bits and pieces of an old 68k design, don't multitask properly, and make it hard allocate more than 64k at a time. Apparently, PalmOS 6 has been released, but neither PalmOne nor Sony are even bothering to make handhelds with it; it's too little too late. The only thing that has kept the platform alive is the fact that there are lots of good applications for it and that kept the original GUI more or less intact.
All Microsoft had to do is show up to the party. WinCE isn't a great operating system, but it's a lot better than PalmOS. The thing that has been holding PocketPC back is its awful UI.
My hope would be that PalmOne (the hardware part of Palm) explores some new ideas: Symbian is a great system they could ship right away, or they could adopt one of the Linux-based PDA environments and port a PalmOS emulation layer to that to run all the Palm legacy applications.
You buy a $50 printer. If you don't agree to the EULA, it's a piece of junk. If you take it back, your retailer is going to keep track of you and make future returns harder, and the time it takes to return it isn't worth the printer anyway.
"TSA is aware of, and sensitive to, the need to preserve Americans' freedom while pursuing better security," the agency wrote. "In implementing a new security measure that affects these interests, it is necessary to move deliberately and cautiously."
Maybe they should be sensitive to the need to preserve everybody's freedom and security, American or not. It's the right thing to do. Even if that doesn't matter to the TSA, tourism figures should and tourists will stop coming if they don't feel welcome.
You might as well say "Any code that requires Intel x86 or a derivative to run on might as well be owned by Intel."
No. First, most open source source code is not processor dependent. Second, there already exist multiple, independent implementations of the x86 architecture, some even in software. I'm typing at one right now.
Third, with processors, we don't have a choice, but with software, we do. We don't have to let it come again to a situation where a single company sets the agenda. And even if the market went in that direction, I would not want that single company to be Sun.
The irony here is that because of Sun's requirement that Java implementations pass their extensive compatibility tests, you have the freedom to choose from a variety of different Java vendors.
Name even just one J2SE implementation that is not a licensed derivative from Sun's source code and that has passed compatibility testing.
But for large number of people who develop for the Java platform, and the huge number of users who want their applications to just work, it's a no-brainer that compatibility and choice are more important.
Yes, just like for Windows: one company makes the platform software and has its compatibility requirements, a lot of companies resell it, and the "huge number of users" gets "compatibility and choice". Basically, Sun is offering us the same deal as Microsoft, only the brand name is a different one. Thanks, but no thanks.
And most of that development was done by Bill Joy, the same guy who wrote the original code in the first place - he just kept on doing what he had been doing, but Sun paid him to do it.
Yes, that's well known. What are you trying to imply?
That's the BSD license, like it or dislike it. Nothing to do with "good" and "evil".
But it has something to do with whether a company is open source friendly. A company that takes a BSD-licensed piece of software and builds a proprietary product around it is not open source friendly. They aren't doing anything illegal, but they have made it pretty clear through their actions where they stand on open source.
Sun took the BSD code developed at Berkeley and built a proprietary system out of it. I think that counts as "being founded on open source", just not in a good way: their source code went from fairly open to quite proprietary.
As for the posters who are claiming that Sun is just another Microsoft and whatnot, just because a company is large and competitive doesn't mean that it's always patently evil.
Sun isn't evil, but they are dangerous. They are dangerous because they still effectively own a widely used platform (Sun Java) while at the same time their business is seriously threatened.
Imagine Java on Linux became the standard way of running web applications and Linux desktop systems; the majority of Linux servers and desktops used Sun Java in some essntial capacity. Then, Sun's hardware business fails and in the turmoil, they get new management. What is new management going to say? They are going to see that Sun is shipping the software that powers Linux servers and Linux desktops comes from Sun and they are going to try to make some money out of that. How? In the next Java release, they offer "Java Lite" and "Java Pro" versions. The "Java Lite" version comes without a JIT, with a lot of functionality disabled, and with a simple garbage collector. If you want the real thing, you just pay $15 for each CPU and you are up and running again. You may cling to your old binaries for a little while, but with the next standard library upgrade, you just don't have a choice anymore but to upgrade. And because there is no open source Java implementation that can really run the complex server and desktop software, you have no choice but to pay.
The only insurance against such scenarios is to have a real open source or free(-dom) implementation of Java.
Sun can contribute millions of LOC to open source projects and still hold on to crucial portions of the code or crucial patents. Take, for example, the Java Desktop System. It may run on top of Linux and be mostly Gnome. Sun might even donate all the Java code that is part of the Java Desktop System under an open source license. But no matter what they do, what makes it work is Sun Java: there is no open source implementation (and, in fact, no implementation that doesn't include some Sun-licensed code) that runs it. Any code that requires Sun Java or a derivative to run on might as well be owned by Sun.
Of course, in the long run, Sun can't prevent Java from being cloned and from losing control over it. No matter how hard Sun fights, uncertified and incompatible open source implementations will become more and more widely used. But Sun sure is fighting hard to stay in control (all in the name of WORA, of course).
(As for Sun's open source record, I think one might call it "mixed"; but that's a separate thread. Just don't automatically take frequent claims from Sun officials that everybody used to agree that they were "the good guys" as truth.)
That's not true. A wired article (that I can never find) state[...]
There are lots of ideas and trials for dealing with nuclear waste, many of them quite plausible. But even a successful trial isn't the same as demonstrating that something is economical and safe.
Place your waste in front of a high energy radiation gun and accelerate the breakdown by many many times. The waste product is heat which can be made into power and the 60/40 split. Probably lead and radon gas.
You can reduce the amount of highly radioactive waste, but you end up with even more radioactive waste that is still very dangerous. Furthermore, it is anything but clear that such an approach would be economical.
Microsoft does not innovate. It copies, assimilates, and polishes. Microsoft software is never groundbreaking in any significant way
OK, we agree so far.
it just is usually more usable, better integrated, more reputable,
No, it is never that either.
and cheaper than the competition.
Now you're getting close. Microsoft undercuts the competition to drive them out of business. The other thing they do is time precisely when they release and what features they include in their releases, again, with an eye towards killing the competition. But such precise timing means that engineering and quality control schedules become secondary to sales and competitive concerns, and it shows in in their products.
Of course, Microsoft is also changing. These days, they are less and less able to do what they used to do in the past. Now that they have millions of whiny enterprise customers, they are in the same boat as IBM once was: a slow-moving dinosaur ready to be hurt badly by the next new fast growing company. Microsoft will meet its own Microsoft soon, and they will likely fall faster, further, and harder than IBM did.
That may not be an argument Google wants to make. A lot of sites on the Internet have their own terms of service that prohibit some of what Google is doing with their data (like copying it over into their own databases and displaying it from their cache). If Google starts making arguments that TOS are enforceable, they may seem them enforced against them, and they have a lot more to lose in the process.
If the design could be made stable enough for longer flights, engineers believe the aerospace and defense industries might one day use it to build cheaper rockets and faster missiles.
Cheaper rockets and faster missiles--I feel safer already. And just think: while it may deliver a deadly nuclear warhead, at least it doesn't pollute on the way to delivering it.
There are two things to consider. One: what happens when an explosion occurs. Two: how likely is an explosion going to occur. Arguably, hydrogen explosions probably tend to be less serious. However, given that hydrogen storage is both more complex and the handling of hydrogen is something new to most people, the likelihood of an explosion seems higher, at least until the technology is more widely used.
Slashdot Mirror
Researchers could freelance and sell to such "brokering" companies, never worrying about products.
I doubt they can freelance and survive on a once-a-month part-time gabfest at Myhrvold's office. That only works because the tax payer and general population is paying most of the cost of generating these ideas in the first place: universities and university salaries.
Having somebody floating around whose only real motivation is ferreting out such scum and getting them to pay for the hard work that they're trying to sponge off of is a good thing.
Have you ever written any substantial piece of code? Chances are that you are infringing dozens of patents. Are you "scum" or a "sponger" because of it? I don't think so, since I doubt you even know of the existence of the patents you are infringing.
Many ideas that are being patented are so obvious that many people have them independently. The one who happens to be first to the patent office wins.
The real scum are the people who patent things that they know full well (or should know) are part of the public domain: ideas that others have talked about, ideas that have been discussed, ideas that are in textbooks. They steal from the public ideas and property to the tune of billions of dollars.
What I find interesting is how far we have come from a real free market. Ask any student of economics... what we have now in America and Europe is not a free market, rather there is government intervention propping up the strong and upsetting the "natural" balance that we would find with free market forces.
Governments need to intervene in markets in order to keep them free: it's only government intervention that can ensure that contracts are enforceable, that companies don't collude, and that people get the information they need.
The problem is that with the power to intervene for the purpose of keeping markets free, governments are also tempted to intervene for the purpose of lining the pockets of their private supporters.
Patents used to require hard work, experimentation, and demonstrating that something actually works. These days, Nathan invites some buddies for a "gabfest" and writes up the results as dozens of patents. Nobody cares whether it actually works, it just needs to contain a high enough proportion of patents that can be used to pressure companies to pay in order to pay for itself and make a handsome profit. And for a few thousand dollars a pop paid to his buddies for ideas that people can generate in a few minutes, Nathan's company gets to control millions of dollars of investments.
Unfortunately, many of the usual suspects are a member of this little club. Fortunately, if this company can do it, lots of other companies can do it, too. And if one of their client is being sued by another patents-only company, it doesn't matter how big their portfolio is: cross-licensing won't get them out of their legal troubles.
It seems Opera is growing, and they are doing it by aggressively promoting their products, even goes as far as teaming up with IBM's ViaVoice to allow users execute commands by talking to their computers. These are licensed-features that free browsers will find it hard to justify paying for.
The reason open source browsers don't have voice-guided browsing is because it's a useless gimmick. If there was any kind of demand for it, there are multiple open source speech recognition systems that it could be based on; nothing to license.
Are there any other "licensed-features" you can think of?
If you want to climb on a horse over the implementation of the interface, remember that javascript works on more than IE alone.
I'm not "climbing on horses", I'm having problems with Gmail on many of the devices and browsers that I use.
Most 'simpler' handhelds do Pop3 nowadays so they are/will be covered.
You want to look at a 500M mailbox with a handheld over POP3? I think you are a bit naive.
Furthermore, if they would like to get on the Geeks good side they should make a XUL version of their interface. That would create the killer-app for XUL in one go.
Yes, it would be the "geeky" thing to do. It just wouldn't be the sensible thing to do.
Apple's Spotlight isn't technology, it's a software product. The technology itself is various indexing, search, and user interface techniques that have been around for many years.
The biggest problem with the Gmail interface is that it uses so much JavaScript; it doesn't run at all on many handheld browsers. I think at least they need a simplified XHTML interface.
The POP feature also makes no sense to me; it basically begs you to download messages from their servers when their stated goal is to collect lots of mail. If they offered IMAP access instead, people could keep their messages on the server. They could even use IMAP for placing subtle adds (e.g., message "1" is always some kind of simple ad, but unlike spam or hotmail, there would always only be one advertising message).
An area where Gmail could really do something better is passwords: they really should offer one-time passwords for travelers. Right now, when traveling, there is a high chance with web-based mail that your password gets compromised.
In any case, for fairly little money, you can get large mailboxes with IMAP interfaces from other companies, and you get a lot more control over them than with Gmail. Currently, Gmail's "free" isn't good enough for me to save the money I get with a commercial provider.
You missed the entire point that I was trying to make. Yes, "most people" will like PalmOS better because it has a better UI. But the system software that underlies that UI is horrid: the limitations and differences from commonly used operating systems mean that porting and developing for it is a lot of work and usually requires starting from scratch.
Palm has been able to get by on their better UI for years, but the poor job they have been doing on the underlying system software is catching up with them. And with 640x480 PDAs becoming more common, PPC's poor UI matters less and less anyway.
I don't see GPL violations becoming a big problem. First of all, the loss to the open source community from abuses of the GPL is mainly the lack of contributions. That is partially balanced out by the fact that the company in question is at least not using some proprietary solution. Second, the problem is self-limiting: if a company gets to be large enough, someone is going to notice the GPL violation and the company will likely settle the resulting legal action quickly.
For BSD-licensed software (are there any BSD VoIP solutions?), companies are even welcome to use the software and make it proprietary. Proponents of BSD probably believe that it's better to have companies use open-source derived software even if they don't give back than to lose more companies to proprietary vendors. (Personally, I think it needs to be looked at on a case-by-case basis.)
This one Palm did to themselves. Palm had half a dozen years to turn PalmOS into a modern, reliable 32bit operating system, instead they are still shipping handhelds that emulate bits and pieces of an old 68k design, don't multitask properly, and make it hard allocate more than 64k at a time. Apparently, PalmOS 6 has been released, but neither PalmOne nor Sony are even bothering to make handhelds with it; it's too little too late. The only thing that has kept the platform alive is the fact that there are lots of good applications for it and that kept the original GUI more or less intact.
All Microsoft had to do is show up to the party. WinCE isn't a great operating system, but it's a lot better than PalmOS. The thing that has been holding PocketPC back is its awful UI.
My hope would be that PalmOne (the hardware part of Palm) explores some new ideas: Symbian is a great system they could ship right away, or they could adopt one of the Linux-based PDA environments and port a PalmOS emulation layer to that to run all the Palm legacy applications.
You buy a $50 printer. If you don't agree to the EULA, it's a piece of junk. If you take it back, your retailer is going to keep track of you and make future returns harder, and the time it takes to return it isn't worth the printer anyway.
The lesson? Just don't buy Lexmark.
"TSA is aware of, and sensitive to, the need to preserve Americans' freedom while pursuing better security," the agency wrote. "In implementing a new security measure that affects these interests, it is necessary to move deliberately and cautiously."
Maybe they should be sensitive to the need to preserve everybody's freedom and security, American or not. It's the right thing to do. Even if that doesn't matter to the TSA, tourism figures should and tourists will stop coming if they don't feel welcome.
You might as well say "Any code that requires Intel x86 or a derivative to run on might as well be owned by Intel."
No. First, most open source source code is not processor dependent. Second, there already exist multiple, independent implementations of the x86 architecture, some even in software. I'm typing at one right now.
Third, with processors, we don't have a choice, but with software, we do. We don't have to let it come again to a situation where a single company sets the agenda. And even if the market went in that direction, I would not want that single company to be Sun.
The irony here is that because of Sun's requirement that Java implementations pass their extensive compatibility tests, you have the freedom to choose from a variety of different Java vendors.
Name even just one J2SE implementation that is not a licensed derivative from Sun's source code and that has passed compatibility testing.
But for large number of people who develop for the Java platform, and the huge number of users who want their applications to just work, it's a no-brainer that compatibility and choice are more important.
Yes, just like for Windows: one company makes the platform software and has its compatibility requirements, a lot of companies resell it, and the "huge number of users" gets "compatibility and choice". Basically, Sun is offering us the same deal as Microsoft, only the brand name is a different one. Thanks, but no thanks.
And most of that development was done by Bill Joy, the same guy who wrote the original code in the first place - he just kept on doing what he had been doing, but Sun paid him to do it.
Yes, that's well known. What are you trying to imply?
That's the BSD license, like it or dislike it. Nothing to do with "good" and "evil".
But it has something to do with whether a company is open source friendly. A company that takes a BSD-licensed piece of software and builds a proprietary product around it is not open source friendly. They aren't doing anything illegal, but they have made it pretty clear through their actions where they stand on open source.
Sun was co-founded by Bill Joy, the original author of BSD
Yes. More importantly, he took the BSD code with him and made it proprietary. That's how Sun was founded.
they have given the community Java,
They haven't "given" Java, they still own it.
Open Office,
Yes, that's nice. They didn't write it and it's a bit flaky, but it's still nice.
NFS, & RPC
Junk. And even then, they only "gave it" once it was clear it wasn't worth much anymore.
While Java is not strictly open source it is widely used.
So is Windows.
Sun's John Bosak created XML.
Junk. Just remember who to blame for it.
Sun took the BSD code developed at Berkeley and built a proprietary system out of it. I think that counts as "being founded on open source", just not in a good way: their source code went from fairly open to quite proprietary.
As for the posters who are claiming that Sun is just another Microsoft and whatnot, just because a company is large and competitive doesn't mean that it's always patently evil.
Sun isn't evil, but they are dangerous. They are dangerous because they still effectively own a widely used platform (Sun Java) while at the same time their business is seriously threatened.
Imagine Java on Linux became the standard way of running web applications and Linux desktop systems; the majority of Linux servers and desktops used Sun Java in some essntial capacity. Then, Sun's hardware business fails and in the turmoil, they get new management. What is new management going to say? They are going to see that Sun is shipping the software that powers Linux servers and Linux desktops comes from Sun and they are going to try to make some money out of that. How? In the next Java release, they offer "Java Lite" and "Java Pro" versions. The "Java Lite" version comes without a JIT, with a lot of functionality disabled, and with a simple garbage collector. If you want the real thing, you just pay $15 for each CPU and you are up and running again. You may cling to your old binaries for a little while, but with the next standard library upgrade, you just don't have a choice anymore but to upgrade. And because there is no open source Java implementation that can really run the complex server and desktop software, you have no choice but to pay.
The only insurance against such scenarios is to have a real open source or free(-dom) implementation of Java.
Sun can contribute millions of LOC to open source projects and still hold on to crucial portions of the code or crucial patents. Take, for example, the Java Desktop System. It may run on top of Linux and be mostly Gnome. Sun might even donate all the Java code that is part of the Java Desktop System under an open source license. But no matter what they do, what makes it work is Sun Java: there is no open source implementation (and, in fact, no implementation that doesn't include some Sun-licensed code) that runs it. Any code that requires Sun Java or a derivative to run on might as well be owned by Sun.
Of course, in the long run, Sun can't prevent Java from being cloned and from losing control over it. No matter how hard Sun fights, uncertified and incompatible open source implementations will become more and more widely used. But Sun sure is fighting hard to stay in control (all in the name of WORA, of course).
(As for Sun's open source record, I think one might call it "mixed"; but that's a separate thread. Just don't automatically take frequent claims from Sun officials that everybody used to agree that they were "the good guys" as truth.)
That's not true. A wired article (that I can never find) state[...]
There are lots of ideas and trials for dealing with nuclear waste, many of them quite plausible. But even a successful trial isn't the same as demonstrating that something is economical and safe.
Place your waste in front of a high energy radiation gun and accelerate the breakdown by many many times. The waste product is heat which can be made into power and the 60/40 split. Probably lead and radon gas.
You can reduce the amount of highly radioactive waste, but you end up with even more radioactive waste that is still very dangerous. Furthermore, it is anything but clear that such an approach would be economical.
Microsoft does not innovate. It copies, assimilates, and polishes. Microsoft software is never groundbreaking in any significant way
OK, we agree so far.
it just is usually more usable, better integrated, more reputable,
No, it is never that either.
and cheaper than the competition.
Now you're getting close. Microsoft undercuts the competition to drive them out of business. The other thing they do is time precisely when they release and what features they include in their releases, again, with an eye towards killing the competition. But such precise timing means that engineering and quality control schedules become secondary to sales and competitive concerns, and it shows in in their products.
Of course, Microsoft is also changing. These days, they are less and less able to do what they used to do in the past. Now that they have millions of whiny enterprise customers, they are in the same boat as IBM once was: a slow-moving dinosaur ready to be hurt badly by the next new fast growing company. Microsoft will meet its own Microsoft soon, and they will likely fall faster, further, and harder than IBM did.
That may not be an argument Google wants to make. A lot of sites on the Internet have their own terms of service that prohibit some of what Google is doing with their data (like copying it over into their own databases and displaying it from their cache). If Google starts making arguments that TOS are enforceable, they may seem them enforced against them, and they have a lot more to lose in the process.
If the design could be made stable enough for longer flights, engineers believe the aerospace and defense industries might one day use it to build cheaper rockets and faster missiles.
Cheaper rockets and faster missiles--I feel safer already. And just think: while it may deliver a deadly nuclear warhead, at least it doesn't pollute on the way to delivering it.
And why should exhaling air change the weight of the corpse at all?
There are two things to consider. One: what happens when an explosion occurs. Two: how likely is an explosion going to occur. Arguably, hydrogen explosions probably tend to be less serious. However, given that hydrogen storage is both more complex and the handling of hydrogen is something new to most people, the likelihood of an explosion seems higher, at least until the technology is more widely used.