For many months after Hwang's 2004 publication, rumors had spread in scientific circles that the eggs Hwang used to achieve that landmark result had been taken from a junior scientist in his lab. That situation, if true, would be in violation of widely held ethics principles that preclude people in positions of authority from accepting egg donations from underlings. The rules are meant to prevent subtle -- or not-so-subtle -- acts of coercion.
Questions have also circulated as to whether the woman received illegal payments for her role.
Schatten said that Hwang had repeatedly denied the rumor and that he had believed Hwang until yesterday. "I now have information that leads me to believe he had misled me," Schatten said. "My trust has been shaken. I am sick at heart. I am not going to be able to collaborate with Woo Suk."
On the other hand, if you are over 25, just work on your resume. If you've made it this far without the degree, it's not going to help you climb the wage ladder.
I'll have to agree with this one. I wasn't getting too far in the job hunt until I added some bullets to my resume saying I had invented the CD-ROM, the web browser, and Java. That was a lot more effective than that Ph.D. I bought^H^H^H^H^Hearned from WorldWide University.
So, with Dracula, for instance, Bram Stoker held the textual copyright, which expired decades ago--but the publishing company (Penguin, in this case) owns the typographical copyright.
To answer what I think your question is, no, other countries will not honor the patent. But that is true for any patent, simply because the patent system is only national in scope - that is, U.S. patents only apply in the U.S., just like U.K. patents only apply in the U.K., Japanese patents only in Japan, etc.
This is one way that patents differ from copyright - because of the Berne Convention, copyright granted in one country automatically applies to (almost) every other country in the world. But patents have to be applied for on a nation-by-nation basis - well, for the most part, though there is a European Patent Convention which in effect allows applying for patents in most of the countries of Europe at once.
No, actually, the government does have automatic license (permission to use) in the legal sense. I think what you mean is the government doesn't have "free" (as in beer) license to the patent.
Also, if you compare the remedies available for claims against government infringement with those available for non-government infringement (found in United States Code Title 35 Part II Chapter 29), the government does get the better part of the deal. In particular, a patent claimant cannot get an injunction to stop the government from using a patent.
From what I recall, part of the reasoning behind this distinction goes back to World War I, where the government wanted to make sure patent claims could not be used to stop the development and production of new weapons.
See United States Code Title 28, Part IV, Chapter 91, Section 1498. This actually covers the claim procedures when the government makes use of a patent, but if you use this to key some Google searches, you'll get plenty of background on what it is about.
IANAL either, but that's not quite right. Under U.S. law, the Federal government can make use of any patent without first negotiating a license. However, the government must still pay appropriate compensation. Disputes about what is "appropriate" go to the U.S. Court of Claims.
Essentially, it amounts to the "intellectual property" analogy of eminent domain. And just think of the political flamewars you can start by pulling those two concepts together!
The answer, of course, is that they don't all come from Nigeria. In the 90(!) of them I have received in the last week, they've come from the Netherlands, the UK, Palestine (via France and Ireland), Burkina-Faso, the UAE, China, Iraq (via Switzerland), the Philippines, Cote d'Ivoire [Ivory Coast], Benin, Zimbabwe, South Africa, Spain, India, Liberia, the Congo, Russia, Togo, Swaziland, Ghana, Hong Kong, Taiwan, Senegal, and a couple labeled vaguely "West Africa" or just "Africa." Oh, and a passel of them from Nigeria, of course. [I'm just reporting the nominal origins of the mail here; often this is faked, of course. In fact, a lot of the allegedly Nigerian mail really comes from South Africa or other places.]
Looking over this crop, it does appear Nigeria still has first place, but the UK, the Netherlands, and Russia in particular are moving up fast. So it's not just a third-world thing.
To answer your other question, I guess scammers in the US are too busy phishing to bother with relatively labor-intensive deals like the typical Nigerian scam.
And in case you're wondering, yes, I do have a job where my email address has to be prominently posted on a number of web pages. And yes, I do have 3 layers of spam filtering on the account - these all came out of the spam trap.
We tried setting up a cluster of a dozen 1U rackmounted servers a few years ago and had similar problems. We added internal fans, etc., but still ended up losing around half of them to heat issues. Part of the problem is when you have a stack of identical 1U servers, the processors essentially sit on top of each other, so effectively heat each other up. As I recall, it was nodes 2-7 - the ones in the upper part of the stack, but not all the way to the top - which had it worst and died first.
Since then, we've found rack fans that help a whole lot - mount one at the bottom to pull cool air into the rack, and then ones in the back and on top to pull warm air out. If you have a big stack, there are also other fan units you can insert in the middle to cut down on heat buildup in the stack. These units are not cheap - running in the hundreds of dollars each - but beat the price of replacing dead systems.
The other thing we did was give up on 1U servers - we don't use anything smaller than 2U anymore, and have had no comparable problems since. I think the 1U size is just too narrow for effective air circulation. If you think about it, it's not much thicker than a laptop, so unless you use laptop-type components in it, you've got to expect heat issues.
Here's
a link to the original discussion on Slashdot. If you google for "FrontPage disparages Microsoft" (without the quotes), you'll find plenty of sites discussing the issue, including scans of the license in question. (It was only in a printed insert and not online at the Microsoft site.)
As written, the EULA did say "You may not use the Software in connection with any site that disparages Microsoft, MSN, MSNBC, Expedia, or their products or services...." which certainly sounds bad enough.
Probably the intent was that it only apply to things like Microsoft logos included with the software (which is what current FrontPage EULAs
say), but the license is so poorly written it's impossible to say definitively what "the Software" covers.
So a person with a PHD in Astrophysics would get a 5% credit card and someone with just a BA in the same field would get 15%.
Considering the job and earning prospects of the typical astrophysics Ph.D., I think you might have that backwards. (The BA could at least get a job as a programmer like everybody else.)
What we're talking about is taking your whole foundation and throwing it out the window i.e Netscape's browser or Microsofts Win32 API.
Or throwing out the original Windows core in favor of NT? (Yes, I know about NT's "spiritual" heritage from the next generation VMS project, but the core code itself is about as complete a rewrite from any predecessor as you'll ever see.)
But there's a big difference between rewriting parts and rewriting your complete software base.
Sure, and obviously it's rarely worth the trauma of the latter. (Especially when change has become as traumatic as it appears it has at Microsoft.) But where would Microsoft be now if they hadn't made the switch?
I still don't see your point. Your saying that it was better to write code for a platform people could afford. Wow. Considered going into marketing? What this has to do with starting a new codebase, I don't understand. Their choice to "abandon" Multics wasn't voluntarily and the fact they got a cheap platform for UNIX was because they didn't get any funding.
Now you're being deliberately obtuse. Obviously, leaving Multics and rewriting from scratch was not Ritchie, et al.'s first choice. But because they were forced into it, and because, partly out of necessity and partly out of inclination, they threw out inessential elements, they ended up with a system far more portable and usable than Multics ever was. Yes, Multics had many innovative concepts - the problem was, it had too many of them, including many of questionable value.
Have you ever seen the original Multics architecture document? It is far longer than a complete printout of the main documentation and source code for the Unix Version 6 kernel and principal utilities (probably later versions as well, but those are the ones I saw placed side by side). If something on the scale of that conglomeration had been what Bell Labs was releasing to the world, it would never have been picked up.
I mean, yes, there is code that's beyond saving. But in a professional company developing software for a living? Then something is fundamentally wrong and no rewriting will help you. In the case of a normal company, developing shrinkwrap software, abandoning a codebase is moronic.
Then I guess most major software companies have been moronic at some point or other. (OK, I'll grant that that's probably true on other grounds.) Take even Microsoft as an example - what legacy of the original PIA exists in NTFS? It was essentially rewritten at one point when it was decided the original wasn't maintainable.
I honestly don't see your point here? An ill-backed system doesn't leave a legacy. Gee whiz, I'm suprised. The point is that if they hadn't rewrote it from scratch (and as we should note, they didn't want to either) we might have ended up with a mature "UNIX" much quicker.
You're not getting it. It was in large part because they had to rewrite Unix from scratch - for whatever relatively cheap, accessible hardware they could come up with - that it became as successful as it did. "Support" had nothing to do with it - after all, during most of the '70's, Unix was essentially completely unsupported. Universities and other places picked up Unix because it was tractable and ran on affordable hardware. If it had been a more complete Multics clone, I doubt it would have had anywhere near the same level of acceptance at the time.
The thing, that a lot of people seem to fail to grasp here, is that Netscape (the company) died because of this decision.
Actually, they were already pretty much dead at the time. Perhaps the code rewrite, as expensive and time-consuming as it was, was the only way for anything to survive out of it.
And if your code is such as mess that it doesn't encourage contributions: refactor and clean up, don't throw the baby out with the bath water.
Yeah, sometimes this is correct, but sometimes it isn't. For two more (ancient) examples, consider troff and TeX. Ossanna's original version of troff was written mostly in fairly impenetrable assembler; after his untimely death in a car crash, it sat unmodified for years until someone reimplemented it more or less from scratch from his (fortunately quite detailed) specifications. By contrast, modern (so to speak) versions of TeX still show a clear legacy of Knuth's original code; there simply has been no need to do a complete rewrite.
A dogmatic assertion that rewrite is never indicated is just as wrong as one that it's always a good idea.
Multics took a hard hit, when early on BTL dropped out of the project, which it never recovered from.
And yet it was released, during the long period that AT&T couldn't quite figure out what to do with Unix. But for all of that, it seems to have left little legacy today, beyond those features adopted by Unix. The only exception I can think of offhand are ACLs, which were not in the original Unix systems. Of course, part of the reason they exist today is they were picked up by later versions of Unix.... (Paging, dynamic linking and SMP support are of course other examples of things later picked up by Unix, but the Unix versions aren't really direct Multics inheritances in any meaningful sense.)
Even though they did this, they imported many ideas and experiences from Multics.
And discarded just as many. That is exactly my point. Ritchie and Thompson and the rest of the group were successful because they exercised good judgment in what to retain (e.g., the directory structure) and what to discard (e.g., the segmented memory model).
Or we would have had a Firefox 3 years ago.
Or perhaps never at all. A large body of pre-existing code can be a disadvantage, if it is not amenable to adaptation for new purposes, and is not organized in a way to encourage new contributions. This is obviously true for open-source projects, but applies to proprietary software as well.
A silly example is provided by the original Unix Bourne shell. Bourne's source code was written (using preprocessor macros) in a weird Algol-like format, which made others reluctant to do anything with it. It was only when it was rewritten by Korn and others later that it became a subject of active development again.
It contains every little obscure bugged that grandma Uxbuklu in outer Mongolia have ever encountered.
Of course, it also contains every undiscovered little obscure bug left by a programmer who has since moved to Outer Mongolia...
I would recommend you read what Joel has to say, since he say it so much better than I have time to do.
Joel generally has interesting things to say, but I think this is one case where he's a bit off the mark. Perhaps some code he wrote at Microsoft got trashed, and he's been resenting it ever since.
The fact of the matter is, there are problems either way, in keeping old code or discarding it. I don't think a default impulse one way or the other is helpful - you really need judgment to determine what to do.
As an example, was Dennis Ritchie wrong to toss away all that mass of Multics code and write Unix from scratch? Somehow, Multics didn't seem to benefit in the long run from its "head start."
Or, in the case Joel harps on, was it really the wrong decision for the Mozilla crew to toss the old Netscape code and rewrite Gecko from scratch? Perhaps at the time Joel wrote his article it wasn't clear that they had taken the correct approach, but I'd say events have shown the Mozilla team's decision was correct.
Actually, if you check later on in the paper, they test a Dell Latitude C810 laptop as well. And in fact they find (section 7) that their techniques don't work so well there - clock skew varies depending on whether the laptop is on battery or line power, and in the latter case whether the battery is charging or not. Of course, anyone who's ever run adjtimex -c on a laptop has seen this....
I have found that Bayesian filtering is essentially 100% effective on 419 scam mail. As is obvious when reading any of them, they have a very distinctive vocabulary...
The "trick," such as it is, is to maintain three separate Bayes databases - a "good" one, a "spam" one, and a "419" one. Filter with good vs. spam first, and then with good vs. 419. This seems to work better than just lumping 419 mail in with other spam, since as Quinlan notes, the 419 scam mail tends to have little content in common with other spam. But with a separate filter, it can be identified with essentially 100% accuracy.
Well, three, eight, and six consecutively is a pretty distinct pattern. I don't know about you tin-foil hatters out there, but '386' is unique enough to claim as a trademark.
Uh, no, it's not. The decision was that you couldn't trademark a number. That's why Intel went with "Pentium" rather than "586."
Slashdot Mirror
You're right, you shouldn't worry about that rootkit of yours. I can take care of it from here. Anything else you want me to remove while I'm at it?
Here's a snippet of the relevant section:
That's way closer to the average CIA analyst's daily job than you might realize....
This is one way that patents differ from copyright - because of the Berne Convention, copyright granted in one country automatically applies to (almost) every other country in the world. But patents have to be applied for on a nation-by-nation basis - well, for the most part, though there is a European Patent Convention which in effect allows applying for patents in most of the countries of Europe at once.
Also, if you compare the remedies available for claims against government infringement with those available for non-government infringement (found in United States Code Title 35 Part II Chapter 29), the government does get the better part of the deal. In particular, a patent claimant cannot get an injunction to stop the government from using a patent.
From what I recall, part of the reasoning behind this distinction goes back to World War I, where the government wanted to make sure patent claims could not be used to stop the development and production of new weapons.
See United States Code Title 28, Part IV, Chapter 91, Section 1498. This actually covers the claim procedures when the government makes use of a patent, but if you use this to key some Google searches, you'll get plenty of background on what it is about.
Essentially, it amounts to the "intellectual property" analogy of eminent domain. And just think of the political flamewars you can start by pulling those two concepts together!
Looking over this crop, it does appear Nigeria still has first place, but the UK, the Netherlands, and Russia in particular are moving up fast. So it's not just a third-world thing.
To answer your other question, I guess scammers in the US are too busy phishing to bother with relatively labor-intensive deals like the typical Nigerian scam.
And in case you're wondering, yes, I do have a job where my email address has to be prominently posted on a number of web pages. And yes, I do have 3 layers of spam filtering on the account - these all came out of the spam trap.
Since then, we've found rack fans that help a whole lot - mount one at the bottom to pull cool air into the rack, and then ones in the back and on top to pull warm air out. If you have a big stack, there are also other fan units you can insert in the middle to cut down on heat buildup in the stack. These units are not cheap - running in the hundreds of dollars each - but beat the price of replacing dead systems.
The other thing we did was give up on 1U servers - we don't use anything smaller than 2U anymore, and have had no comparable problems since. I think the 1U size is just too narrow for effective air circulation. If you think about it, it's not much thicker than a laptop, so unless you use laptop-type components in it, you've got to expect heat issues.
As written, the EULA did say "You may not use the Software in connection with any site that disparages Microsoft, MSN, MSNBC, Expedia, or their products or services...." which certainly sounds bad enough.
Probably the intent was that it only apply to things like Microsoft logos included with the software (which is what current FrontPage EULAs say), but the license is so poorly written it's impossible to say definitively what "the Software" covers.
(OK, so it's not exactly an "accent," but you have to admit he can sound "strange" at times.)
Have you ever seen the original Multics architecture document? It is far longer than a complete printout of the main documentation and source code for the Unix Version 6 kernel and principal utilities (probably later versions as well, but those are the ones I saw placed side by side). If something on the scale of that conglomeration had been what Bell Labs was releasing to the world, it would never have been picked up.
Then I guess most major software companies have been moronic at some point or other. (OK, I'll grant that that's probably true on other grounds.) Take even Microsoft as an example - what legacy of the original PIA exists in NTFS? It was essentially rewritten at one point when it was decided the original wasn't maintainable.A dogmatic assertion that rewrite is never indicated is just as wrong as one that it's always a good idea.
A silly example is provided by the original Unix Bourne shell. Bourne's source code was written (using preprocessor macros) in a weird Algol-like format, which made others reluctant to do anything with it. It was only when it was rewritten by Korn and others later that it became a subject of active development again.
The fact of the matter is, there are problems either way, in keeping old code or discarding it. I don't think a default impulse one way or the other is helpful - you really need judgment to determine what to do.
As an example, was Dennis Ritchie wrong to toss away all that mass of Multics code and write Unix from scratch? Somehow, Multics didn't seem to benefit in the long run from its "head start."
Or, in the case Joel harps on, was it really the wrong decision for the Mozilla crew to toss the old Netscape code and rewrite Gecko from scratch? Perhaps at the time Joel wrote his article it wasn't clear that they had taken the correct approach, but I'd say events have shown the Mozilla team's decision was correct.
Actually, if you check later on in the paper, they test a Dell Latitude C810 laptop as well. And in fact they find (section 7) that their techniques don't work so well there - clock skew varies depending on whether the laptop is on battery or line power, and in the latter case whether the battery is charging or not. Of course, anyone who's ever run adjtimex -c on a laptop has seen this....
The "trick," such as it is, is to maintain three separate Bayes databases - a "good" one, a "spam" one, and a "419" one. Filter with good vs. spam first, and then with good vs. 419. This seems to work better than just lumping 419 mail in with other spam, since as Quinlan notes, the 419 scam mail tends to have little content in common with other spam. But with a separate filter, it can be identified with essentially 100% accuracy.