This bill (referred to S.877, even by the Clerk of the House) hasn't actually passed the House yet. The House is still in session, at 2:30 AM. There was a voice vote, but it wasn't decisive, and a roll call vote was scheduled. To save time, all the roll call votes today will be run at the end of the "day". The roll call vote is on the calendar, but it hasn't happened yet. At this moment, the House is voting on whether to recommit the Medicare prescription drug benefit bill back to committee.
This bill could still die. Call your Congressional office. The staff is still there, very tired, and answering the phone.
I just did. The House is in session, debate on S.877 has concluded, and the vote will be some time in the next few hours, in the middle of the night.
Congressional staffs are still in their offices at midnight. Tell them to vote NO on S.877, because it legalizes spam. CALL NOW
First "Hyperthreading", now "prioritized hyperthreading".
It's amusing seeing this. It reflects mostly that Microsoft has finally managed to ship in volume OSs that can do more than one thing at a time.
(Bear in mind that most of Microsoft's installed base is still Windows 95/98/ME. Transitioning the customer base to NT/Win2K/XP has gone much more slowly than planned.)
But Microsoft takes the position that if have multiple CPUs, you have to pay more to run their software. So these strange beasts with multiple decoders sharing ALU resources emerge.
The current trademark status of "Fedora" is that Red Hat has applied for US trademark registration, but their application hasn't even been reviewed yet.
Red Hat has only filed an "intent to use" the trademark; the date "first use in commerce" is "unavailable".
The Fedora Project can now also apply for trademark registration, with a "first use in commerce" date several years old. If they do, the examiner will have to sort this out.
The Fedora Project probably has priority, which is based on the date of first use in commerce.
There's nothing in the mineral-resources area valuable enough to justify going to the moon for it.
Certainly not with any propulsion technology we have today.
Maybe if someday we get fusion or something better. But chemical fuels are far too weak.
We need a better propulsion system.
It's worth noting that we haven't had a new energy source in the last half century.
Sending something to the asteroid belt that finds a small asteroid and boosts it back in our direction might be possible, but it's potentially rather dangerous.
A typical fallacy of Linux developers is that they think that bolting a "front end" onto a command-line program makes it user-friendly. This never works.
When something breaks, which happens frequently, the end user is faced with error messages from a layer they didn't know existed. The "front end" typically has no idea to what to do about such errors. Figuring out what happened in such situations is usually quite difficult, and well beyond the ability of the user receiving the message.
The basic problem is that the UNIX command line interface assumes the input of large numbers of correct commands, and returns very little machine-readable information. That's not a fundamental problem with a command/response interface; consider HTTP, SMTP and FTP, with well-understood numbered return codes. It's a legacy of the UNIX teletype interface. It's a terrible interface for something that resembles a client/server system.
The UNIX crowd will never get this right. It takes a complete rethink, like the original Mac interface or HTTP, to get it right. The UNIX world can't even get rid of all those stupid files in/etc.
Actually, I meant to say hill-climbing rather than gradient ascent. True gradient ascent implies you can evaluate the local gradient, i.e. you can compute the derivatives which represent the local slope direction. You can't do that for discrite problems like this one. Gradient ascent is more appropriate to problems expressed as differential equations. (Although, having struggled with nonlinear differential equation solving in high-dimensional spaces for several years, I can report that it's no panacea.)
By "broad-front search" I mean one where you're searching the surface defined by the objective function from multiple starting points. GAs, neural nets, and simulated annealing all effectively do this.
All three approaches are really special cases of a more general search operation. Because partisans of each approach use such different terminology, (much of it borrowed from biology) it's not obvious that this is the case.
You could probably get equally good results with plain hill-climbing. Turn on all the optimizations. Then turn them off one at a time and see if things get better. Repeat until no improvement is observed.
Any time you consider using a broad-front search algorithm like a GA, a neural net, or simulated annealing, try plain-hill climbing first. If you try any broad-front search, compare it with plain-hill climbing.
Only if the space being searched is dominated by local maxima (but not too many of them) do the broad-front algorithms help. And they're always slower.
If this guy had demonstrated that a GA did better than a plain hill-climber, he'd have an interesting result. But he hasn't demonstrated that.
The trouble with fighting back is that if you do it ineffectively, you get clobbered, and if you do it effectively, you go to jail for assault and battery.
You need legal advice before beating up a bully today. You can't seriously damage them when they're not directly attacking you.
The funny thing about Larry Ellison in Silicon Valley is that he's mostly ignored here. He has the database business, but nobody else in the Valley does much in that area. His ventures into new technologies like thin clients, video streaming, and supercomputers have all been duds. Oracle is viewed as a large but boring enterprise applications company like Computer Associates, SAP, or Automatic Data Processing.
NASA has become pointless. The purpose of the shuttle fleet is to build the ISS. The purpose of the ISS is to develop ways to keep people alive in space long enough to get to Mars. There are no concrete plans to go to Mars. Going there on chemical fuels will never work very well anyway. Give it up.
Turn the shuttles over to the USAF, let them launch one of them out of Vandenberg when they have to, and dump the Government-funded civilian space program.
Further work on space propulsion systems should be moved to the Department of Energy.
Why, for a media event, are those people using CRTs?
Not only are they big and bulky, CRTs shown on TV flicker unless you get everything synched.
Manipulating chess pieces that way is just silly.
Who are those X3D guys, anyway?
Just to confuse everything, there's an X3D consortium. A few years ago, the angle bracket police decided to convert VRML to XML syntax. Unfortunately, nobody cared about VRML by then.
They'd have to apply for a B-1 visa, to do business in the US, but that's routine. It does, though, mean that they get photographed, identified, and put into US Government files.
Remember, they're pretending to offer a legitimate business deal.
I wonder how far you could get them to go. If you sent one of the scammers a plane ticket to the US, would they come? With a little bit of work and a few hundred dollars, you could probably put them in a US jail.
You can't copyright facts in the US. (Although see the draft intellectual property rules of the Free Trade Association of the Americas on database copyright.) So take the info from the ad, and convert it to a table of items and prices.
MSN's default search doesn't normally return huge numbers of hits for anything. Try "Bush", for example. 712 hits on MSN, 24,800,000 hits on Google.
MSN's paid search hits ("featured sites") are clearly pro-Microsoft. It's less clear that the real search results are. MSN's search results aren't as useful as Google's, but that may just be inferior technology. MSN tends to return far more outdated pages.
Slashdot Mirror
This bill could still die. Call your Congressional office. The staff is still there, very tired, and answering the phone.
I just did. The House is in session, debate on S.877 has concluded, and the vote will be some time in the next few hours, in the middle of the night. Congressional staffs are still in their offices at midnight. Tell them to vote NO on S.877, because it legalizes spam. CALL NOW
We were only a few weeks from the day when California's anti-spam law flatly prohibited it.
This is a huge sell-out.
They are going to get sued. Many times. At $1000 per spam.
Apple's market share today is lower than it was the day Gil Amelio left.
It exists. "The Concept of the Corporation", by Alfred P. Sloan. He put together General Motors.
No, it was an AS/400.
It's amusing seeing this. It reflects mostly that Microsoft has finally managed to ship in volume OSs that can do more than one thing at a time. (Bear in mind that most of Microsoft's installed base is still Windows 95/98/ME. Transitioning the customer base to NT/Win2K/XP has gone much more slowly than planned.)
But Microsoft takes the position that if have multiple CPUs, you have to pay more to run their software. So these strange beasts with multiple decoders sharing ALU resources emerge.
Also, you can register trademarks online in a few minutes. If you have a good domain name, register it.
The Fedora Project can now also apply for trademark registration, with a "first use in commerce" date several years old. If they do, the examiner will have to sort this out. The Fedora Project probably has priority, which is based on the date of first use in commerce.
Maybe if someday we get fusion or something better. But chemical fuels are far too weak. We need a better propulsion system. It's worth noting that we haven't had a new energy source in the last half century.
Sending something to the asteroid belt that finds a small asteroid and boosts it back in our direction might be possible, but it's potentially rather dangerous.
When something breaks, which happens frequently, the end user is faced with error messages from a layer they didn't know existed. The "front end" typically has no idea to what to do about such errors. Figuring out what happened in such situations is usually quite difficult, and well beyond the ability of the user receiving the message.
The basic problem is that the UNIX command line interface assumes the input of large numbers of correct commands, and returns very little machine-readable information. That's not a fundamental problem with a command/response interface; consider HTTP, SMTP and FTP, with well-understood numbered return codes. It's a legacy of the UNIX teletype interface. It's a terrible interface for something that resembles a client/server system.
The UNIX crowd will never get this right. It takes a complete rethink, like the original Mac interface or HTTP, to get it right. The UNIX world can't even get rid of all those stupid files in /etc.
By "broad-front search" I mean one where you're searching the surface defined by the objective function from multiple starting points. GAs, neural nets, and simulated annealing all effectively do this. All three approaches are really special cases of a more general search operation. Because partisans of each approach use such different terminology, (much of it borrowed from biology) it's not obvious that this is the case.
We were all dreading having to program that turkey.
Any time you consider using a broad-front search algorithm like a GA, a neural net, or simulated annealing, try plain-hill climbing first. If you try any broad-front search, compare it with plain-hill climbing. Only if the space being searched is dominated by local maxima (but not too many of them) do the broad-front algorithms help. And they're always slower.
If this guy had demonstrated that a GA did better than a plain hill-climber, he'd have an interesting result. But he hasn't demonstrated that.
The trouble with fighting back is that if you do it ineffectively, you get clobbered, and if you do it effectively, you go to jail for assault and battery. You need legal advice before beating up a bully today. You can't seriously damage them when they're not directly attacking you.
The funny thing about Larry Ellison in Silicon Valley is that he's mostly ignored here. He has the database business, but nobody else in the Valley does much in that area. His ventures into new technologies like thin clients, video streaming, and supercomputers have all been duds. Oracle is viewed as a large but boring enterprise applications company like Computer Associates, SAP, or Automatic Data Processing.
Turn the shuttles over to the USAF, let them launch one of them out of Vandenberg when they have to, and dump the Government-funded civilian space program.
Further work on space propulsion systems should be moved to the Department of Energy.
Manipulating chess pieces that way is just silly.
Who are those X3D guys, anyway? Just to confuse everything, there's an X3D consortium. A few years ago, the angle bracket police decided to convert VRML to XML syntax. Unfortunately, nobody cared about VRML by then.
Remember, they're pretending to offer a legitimate business deal.
I wonder how far you could get them to go. If you sent one of the scammers a plane ticket to the US, would they come? With a little bit of work and a few hundred dollars, you could probably put them in a US jail.
You can't copyright facts in the US. (Although see the draft intellectual property rules of the Free Trade Association of the Americas on database copyright.) So take the info from the ad, and convert it to a table of items and prices.
Something like this was tried in Japan about ten years ago, as a dating device. It didn't work.
MSN's paid search hits ("featured sites") are clearly pro-Microsoft. It's less clear that the real search results are. MSN's search results aren't as useful as Google's, but that may just be inferior technology. MSN tends to return far more outdated pages.
Promised for Q4 2003.