The problem is that the capacity to handle trafic on a highway decreases as people slow down. A highway that is near or at capacity will end up at a standstill in places if people slow down at any point.
If your highway is really that full, you need eather 2 more highways of that size or 3 times as may lanes on that highway. 8 each way sounds like a lot but it's no where neer as much as you should have. It should be 24 lanes each way or even better, there should be 2 other highways running parallel to it a few miles on eather side.
Seriously though, these studies simply show that the roads that were expanded were only expaneded a tiny percent of what they needed to be.
If you look at that A&M study carfully you will also notice that the green part of the graph, how much people drive, was completely unaffected by both congestion time and roadway expansion. People go where they need to go regardless of weather you build the road to handle the capacity well.
I've said this in other posts but the thing people don't realize is that when a traffic jam happens on a highway that road is handling many times the traffic it should. Safe travel distance on a highway according to most state agencies is 3 seconds. You should pass a fixed point on the ground 3 seconds after the back of the car in front of you. On a busy highway, there ends up bing more like 1 second or 0.5 seconds between cars. If you want that highway to be safe again you need to more than tripple it's size. Seriously. Adding a lane or two to a full highway is not enough. You end up with.75 seconds between cars instead of.5. The road is still full and you still end up with traffic jams.
I'll say it again:
If you want that highway to be safe again you need to
more than tripple it's size.
The "more than" comes from the fact that you should have a mostly empty passing lane and an entrance/exit lane. Also, as is mentioned in the A&M study, when a road is at capacity people avoid it. They switch to 1) other times, 2) other roads and 3) other modes. When you fix the road, people switch back.
In 1999, the Surface Transportation Policy Project (STPP) analyzed data obtained from the Texas Transportation Institute and concluded Texans don't know how to build roads
> There's gotta be new thinking in people moving...focusing not just on environment, but quality of life and practicality,
Absolutely. Let's not stick our heads in the sand for the next 10 years like we have for the last 10. The answer is to dramaticly increase the size and number of our highways.
We are *not* doomed to having undersized roads. Most highway expansions are done because there is congestion or traffic jams. The problem is that by the time a highway gets to that point it is *way* over capacity and they end up expanding a tiny fraction of what is needed.
What most people don't realize is that when a highway gets to the point of having traffic jams just how far over capacity it really is. Heavily traveled highways should have a spare lane for passing on the left and a spare lane for entering and exiting on the right and middle lanes with enough room so the traffic on them can pass with 3 seconds between the back of one car and the front of the one behind it (This is the distance most states recommend for highway travel). That's about 3.5 seconds per car (from the front of one to the front of the one behind) and 4 seconds per truck.
Using this formula a highway that can carry 1 car per second should be 5-6 lanes. Most highways don't start bunching up and slowing down until they are over 1 car per second per lane (this is an educated guess based on observations of traffic in the Boston and New York area where people travel very close in rush hour). This means they are running at about 3 times what a "safe" highway would carry not counting the entrance and passing lanes. A 2 lane (per side) highway that carries 2 cars per second at peak should really be a 8 lane highway (16 lanes counting both sides). A 4 lane highway that is packed should really be a 14 lane highway (28 counting both sides) or more appropriately, 3 5 lane highways. Instead, projects usually add a paltry lane or two which just makes for a highway that is still over full it just has a little more room between the packed in cars which makes it slightly safer.
The problem with highway expansions is that the scale is way off. They don't try to get back to a safe comfortable traffic pattern, they just try to stop the traffic jams. This is a mistake. It makes for roads that are still dangerous, stressful and just end up needing to be expanded again a few years later.
If this sounds like a lot of expensive roads it's not. The expensive roads are the back roads. Highways are free. Most states pay for the building and maintainance of roads with gas tax. The highways with such high levels of traffic make much more in gas tax money than they cost. Most busy highways carry over one car per second. In contrast, those back roads that carry one car per minute or one car per hour still need to be built and tarred and plowed at great expense yet pull in very little gas tax money. Those are the expensive roads, not the highways.
As far as where to put new highways. Build them in a grid. North south highways intersecting with east west ones. Distance between them should depend on population densities. If there are buildings in the way, knock them down. Really. Knock them down. Don't build elevated highways around them, don't dig underground, take the land using eminent domain, knock the buildings down and put roads there. Then, take all the land next to the highways too and lease it back to people temporarily and kick them off when the buildings get old or it's convenient so you can expand the highway next time. In the end, it's a lot better economically to do that than making tens of thousands of people sit in cars with the engine running for an extra 20 minutes every day. Good efficient transportation makes for a clean happy prosperous city. It's worth it.
One of the interesting things about analyzing traffic using the cars/second/lane method is that you'll notice that as traffic slows down the capacity of the highway gets lower. That's because the time that it takes the car to go by g
> And what's gonna happen when you widen the roads? More people are > going to move out further from the city, so more people are on the > road. It's a self-defeating solution.
This is exactly the mentality I was talking about. The roads are too full so lets do nothing. It's not clever people, it's stupid.
Let's not stick our heads in the sand for the next 10 years like we have for the last 10. To understand why not building roads doesn't fix the capacity issue think about it the other way around. If you get a great job somewhere where you need to commute on a rather busy highway will you turn the job down? What if you have no other job opportunities? Don't kid yourself. Do you think the average person would take a job for $5,000 less per year to take 5 minutes off their commute? Our society is simply to specialized and spread out to not have good roads and it just isn't going to get less specialized or less spread out.
We are *not* doomed to having undersized roads. Most highway expansions are done because there is congestion or traffic jams. The problem is that by the time a highway gets to that point it is *way* over capacity and they end up expanding a tiny fraction of what is needed.
What most people don't realize is that when a highway gets to the point of having traffic jams just how far over capacity it really is. Heavily traveled highways should have a spare lane for passing on the left and a spare lane for entering and exiting on the right and middle lanes with enough room so the traffic on them can pass with 3 seconds between the back of one car and the front of the one behind it (This is the distance most states recommend for highway travel). That's about 3.5 seconds per car (from the front of one to the front of the one behind) and 4 seconds per truck.
Using this formula a highway that can carry 1 car per second should be 5-6 lanes. Most highways don't start bunching up and slowing down until they are over 1 car per second per lane (this is an educated guess based on observations of traffic in the Boston and New York area where people travel very close in rush hour). This means they are running at about 3 times what a "safe" highway would carry not counting the entrance and passing lanes. A 2 lane (per side) highway that carries 2 cars per second at peak should really be a 8 lane highway (16 lanes counting both sides). A 4 lane highway that is packed should really be a 14 lane highway (28 counting both sides) or more appropriately, 3 5 lane highways. Instead, projects usually add a paltry lane or two which just makes for a highway that is still over full it just has a little more room between the packed in cars which makes it slightly safer.
The problem with highway expansions is that the scale is way off. They don't try to get back to a safe comfortable traffic pattern, they just try to stop the traffic jams. This is a mistake. It makes for roads that are still dangerous, stressful and just end up needing to be expanded again a few years later.
If this sounds like a lot of expensive roads think about trying to haul that many people using rails. A typical single track of railroad handles about 1/5th of the traffic of a single highway lane. Things scale much better if you replace an entire highway but even with optimistic estimates the cost would be astronomical. And once you have it all built you need to maintain all the stations, trains, drivers, track and it can't do the anywhere to anywhere that roads can do.
As far as where to put new highways. Build them in a grid. North south highways intersecting with east west ones. Distance between them should depend on population densities. If there are buildings in the way, knock them down. Really. Knock them down. Don't build elevated highways around them, don't dig underground, take the land using eminent domain, knock the buildings down and put roads there. Then, take all the land next to the highways too and lease it back to people temporarily and kick them off when the building
The space argument is silly. Most of the space taken up by roads is taken up by the little roads that go to everyone's door. How are you going to move things into/out of a house unless there is a road to it? Highways take up neglidgable space in comparison to how much traffic they carry. That back road to your door that carrys 1 car per hour, or 1 car per minute is much less efficient use of space and money than those highways that carry 3 cars per second. You want to save money and space, have local parking lots for every neighborhood and tripple the size of the highways. The only thing better than that would be to not have local parking garages, leving the local stuff alone and trippling the size of the highways.
> And what's gonna happen when you widen the roads? More people are > going to move out further from the city, so more people are on the > road. It's a self-defeating solution.
This is exactly the mentality I was talking about. The roads are too full so lets do nothing. It's not clever people, it's stupid.
Let's not stick our heads in the sand for the next 10 years like we have for the last 10. To understand why not building roads doesn't fix the capacity issue think about it the other way around. If you get a great job somewhere where you need to commute on a rather busy highway will you turn the job down? What if you have no other job opportunities? Don't kid yourself. Do you think the average person would take a job for $5,000 less per year to take 5 minutes off their commute? Our society is simply to specialized and spread out to not have good roads and it just isn't going to get less specialized or less spread out.
We are *not* doomed to having undersized roads. Most highway expansions are done because there is congestion or traffic jams. The problem is that by the time a highway gets to that point it is *way* over capacity and they end up expanding a tiny fraction of what is needed.
What most people don't realize is that when a highway gets to the point of having traffic jams just how far over capacity it really is. Heavily traveled highways should have a spare lane for passing on the left and a spare lane for entering and exiting on the right and middle lanes with enough room so the traffic on them can pass with 3 seconds between the back of one car and the front of the one behind it (This is the distance most states recommend for highway travel). That's about 3.5 seconds per car (from the front of one to the front of the one behind) and 4 seconds per truck.
Using this formula a highway that can carry 1 car per second should be 5-6 lanes. Most highways don't start bunching up and slowing down until they are over 1 car per second per lane (this is an educated guess based on observations of traffic in the Boston and New York area where people travel very close in rush hour). This means they are running at about 3 times what a "safe" highway would carry not counting the entrance and passing lanes. A 2 lane (per side) highway that carries 2 cars per second at peak should really be a 8 lane highway (16 lanes counting both sides). A 4 lane highway that is packed should really be a 14 lane highway (28 counting both sides) or more appropriately, 3 5 lane highways. Instead, projects usually add a paltry lane or two which just makes for a highway that is still over full it just has a little more room between the packed in cars which makes it slightly safer.
The problem with highway expansions is that the scale is way off. They don't try to get back to a safe comfortable traffic pattern, they just try to stop the traffic jams. This is a mistake. It makes for roads that are still dangerous, stressful and just end up needing to be expanded again a few years later.
If this sounds like a lot of expensive roads think about trying to haul that many people using rails. A typical single track of railroad handles about 1/5th of the traffic of a single highway lane. Things scale much better if you replace an entire highway but even with optimistic estimates the cost would be astronomical. And once you have it all built you need to maintain all the stations, trains, drivers, track and it can't do the anywhere to anywhere that roads can do.
As far as where to put new highways. Build them in a grid. North south highways intersecting with east west ones. Distance between them should depend on population densities. If there are buildings in the way, knock them down. Really. Knock them down. Don't build elevated highways around them, don't dig underground, take the land using eminent domain, knock the buildings down and put roads there. Then, take all the land next to the highways too and lease it back to people temporarily and
Face it. Most people live in cramped spaces because they have to. All things being equal, they would rather have a yard. Would someone in Manhatten like having a big house and a big back yard to play in. Sure. It just doesn't make sense concidering the land would cost about 40 million.
> The car system is falling apart
The car system works well where it is done right. The problem we are having has nothing to do with highways not scaling well. They scale excellently, much better than rail.
The problem is that as a society, people are more specialized in their jobs and most households have two incomes now. The increased specialization makes commutes longer because we need to travel farther to reach a company that does what we do. Historicly we would just move to where the company is but since there is another specialized income in the household we just have to commute. If one person works an hour away from the other there is going to be at least an hour of commuting between the two. There is no way arround that.
This makes for more cars traveling further every day and traffic increases. This isn't going away. There is no magic bullet here to make roads unneccesary. We just need to make more roads.
Car's aren't perfect but they are the most economically efficient solution for most places. The main problem with cars is that most governments have decided not to improve the roads when improvements are needed. When they do improve them they do stupid things like the Big Dig in Boston. Trains are wildly expensive for anything but the most densely populated cities. Segway's are too slow to handle long distance travel. Cars are versitile, quick, efficient, and do their job well.
The problem with getting rid of cars is that I want a back yard. The bigger the better. Most people don't want to live on top of one another in big buildings with no place for their kids to play. A world without cars is a world where everyone needs to be packed in on top of each other so that mass transit can work. I don't like that idea.
If the roads are too crouded, build bigger roads. It's not a hard conept. Why do people think they're doing something clever by not building roads when they should (I live in New Hampshire, north of Boston where commuting is horrible.) We waste thousands of man-hours of time every day, waste tons of gas, increase pollution and make thousands of peoples lives more stressful. It's not celever!
Let's kick the smart people out after they've helped us out and built relationships with people here so businesses will be forced to higher people who have lived here longer. Absolutely stupid!
We should be doing everything in our power to get smart people to move to the US. It's what has made us strong in the past.
This article was written to try to get people to buy technology training. In reality I expect most companies to be reducing the staff of their IT departments to try to become profitable again.
Computer systems these days are built with more redundant and more powerful components that are easier to maintain. Most companies have switched away from the Windows 9x line which saves a awful lot of time dealing with stupid desktop issues. We have also seen the change to web-based software which is getting cheaper and easier to implement and support every day. It is now possible for a small group (3 people or so) to manage a large pool of inexpensive web-servers (20-50) which supports a huge application used by thousands of people. This model is increasingly being used and it works well and saves bundles of money. This, combined with companies new-found zeal for cost cutting will drive the numbers of IT professionals down in the coming years.
My advice is if you aren't in IT, don't try to get into it unless you are really good with computers. If you are in IT but aren't very good at it, think about finding another career. The future in IT may not be very bright.
And Windows will be so much better without DOS
on
Legacy-Free PCs
·
· Score: 2, Interesting
This is such a fluff article. The basic premise seems to be that as soon as we replace some of the stable, time tested aspects of a computer it will magically get much better. Pure fluff.
Today's computers have almost nothing in them that was available 2 years ago no less twenty. The core of a computer is the north bridge chipset. This is where most of the speed is determined and most of the cost comes from. This is where we have DDR Ram, 533 MHz Front side busses, and AGP 8x. Nothing here remotely resembles a PC from 20 years ago. Sure, computers still have a version of the keyboard port they used 20 years ago. We still use it because it's really good at being a keyboard port.
The PCI section was funny. In one breath the article said that PCI express is an evolution from PCI that is invisible to software. The quote was: "mainly a hardware change that will result in simpler motherboard and peripheral designs". Then 5 lines down the article said that when PCI Express is adopted "a whole new class of PC will emerge." Yea, and that class will be slightly different than the class before just like always.
As far as the claims that the hard drive attachment technology hasn't changed much in the last 20 years it's very hard to find anything in modern IDE that existed back in the PC. The physical signaling is very different, the controller is on the drive now, there is a protocol (ATA) running on top of the bus, the addressing has completely changed. Iâ(TM)d say the biggest change with IDE came back around 1993 when ATA was developed to run on top of it. I am a great fan of SerialATA but it is just an evolutionary change in the physical communications layer. That's one of the best things about it, that it is compatible with the "legacy" architecture and yet the article raves about it and then laughably backes it up by saying that the first serial ATA drive out was "quieter and cooler-running than its classic ATA counterparts" Pure fluff.
As for the floppy, it is certainly time for something to be done about it and yet next to no work has been done on a replacement. The floppy disk is a random read-write bootable removable medium that every PC operating system natively supports. There is no other device that can claim that. CD Burners should have replaced the floppy years ago but the manufacturers never got together and built a new standardized low-level interface. Even bootable CD's still emulate a floppy disk and the boot image is limited to the size of a 2.88 MB floppy. The floppy replacement is an issue that now *needs* to be addressed and yet the articleâ(TM)s suggestion is to simply leave it out without anything to replace it's unique functionality.
Every once in a while these fluff articles pop up. "Soon computers will be as simple, cheap and as easy to use as your phone" they spout "and all they need to do is leave out all that old stuff that you don't really need". The thing they seem to miss is that it has already happened. You can go down to the store and buy a nice legacy-free computer with none of those useless 20 year old keyboard ports or dumb serial ports and it's cheap and easy to use and it's a palm pilot and it sucks for doing what computers are good at. There are all kinds of "legacy-free" computers out there, Ipaq, Tivo, smart phones, there's even those super-cool 3com Audreyâ(TM)s that are all the rage because they are legacy-free... except they aren't all the rage because they suck and that's in no small part due to being legacy-free.
Legacy free usually means not compatible with the old stuff and for a computer that means it's less flexible and thus less powerfull and less desirable. There is a *huge* amount of effort that has gone into designing and supporting these "legacy" systems and to suggest that because it's old it should go is to forget a fundamental truth in the PC industry:
If it has lasted this long, it is probably better at some aspect of it's job than anything else and there is worth
> At a conservative estimate, MySQL is 25 years behind the state of the art.
But state of the art isn't what's important with OSS. OSS is about the commodity market and relational databases *are* a commodity now. True, not all the features you need are in both of these databases and it's easy to come up with a feature list where Oracle looks great. That's not the point. When you need a database for a project odds are one of these two (PostgreSQL or MySQL) will give you what you need. A lot of programmers don't like the idea of learning how to code against these two because they already know Oracle and with that knowledge they can tackle any project. The problem is the guy next to you who knows these two OSS databases can tackle that project for $10K less. Who's more valuable as a programmer? The answer isn't always going to be the OSS programmer but it will more often than not.
These databases are especially important for commercial applications that need a DB back end. It's one thing to have a $2-$10K db license that you can share among multiple applications but when you are selling a product that relies on a database backend, using an OSS DB can save you the cost of the comercial database hundreds or thousands of times over.
Ellison will still make money for years to come just like Apple has and Microsoft will. Being relegated to the non-mainstream doesn't mean death, it just means there is a new set of rules. No more rolling in wheelbarrows full of cash by charging big money for commodity software.
Wow! 3DFx was an impressive company with really good technology. I was a NVidia man my self but up until the end I debated the wisdom of my choice. 3DFx had sound technology, loyal customers, and some really powerful IP (their multi-chip rendering was by far the best.) It takes a real moron to run that strong of a company into the ground. It would seem from the outside that they could have backed off, sured things up and lived to see another day. I can't believe someone else hired him! What the hell are these people thinking?
I chose ReplayTV over Tivo a long time ago. I'm a big Linux fan but I just couldn't get over the 10/100 ethernet port and all the cool options it represented. They used http for interaction with the box which is wonderfully flexible. It was a nice modular architecture with lots of expantion possibilities. And months went by and nothing happened. And more months and more nothing. I haven't had a software update on the box for almost a year and there are issues that need to be fixed. The company has been complietely impotent.
In my experience, there is a type of person who loves to be in charge and should never be allowed to do so because they suck at it. They see most successes as their's and most failures as someone else's. To them, every prolem has an easy solution and it's the best one. I have been blessed to work with some great people who know how to encourage, lead by example, trust in the wisdom of the people who know what they are talking about, and stick to what they do best.
You can blame these things on bad engineers, small budgets and other logistical issues but let's face it: A good leader *makes* things happen. An average leader *let's* things happen. It takes a bad leader to stop things from happening and I get the feeling that's the type of leader Greg Ballard is.
Of course Java can call native libraries, that was part of my argument too. I don't hate Java, it's just not appropriate for libraries that you want to share between languages. C is the only option there which is why SSL support, rsync support, DB interfaces, OpenGL etc are all written in C. Things written in C can be used in any language. You can't say that about any other language.
A good example of this is ImageMagick. It's a nice library for manipulating images. It's written in C but you can use it from Perl, Python, PHP, C, C++ and Java. If it was written in any other language it would only be usable in that language.
C is prone to memory issues because the standard way it deals with strings is horribly dangerous. Strings shouldn't be represented by a simple pointer to a place in memory. Strings should, at the very least, be structurs with information on length of buffer space and length of the currently stored string. All functions that operate on strings should gracefully handle trying to read or write past the end. Many libraries have already implemented passing a length in with a string when one is passed in but this is not done in a standardized way. We need to get rid of null terminated strings in C.
And putting on my Perl advocate's hat, I doubt it's easier to invoke C from Java than Perl. I've heard that Perl has the tightest C integration of any modern portable language. Many simple C functions are directly accessible from perl (printf, sprintf, select). Most are available by using one of the standard modules included in the Perl 5 distribution. More advanced things like SSL, Rsync, ImageMagik, and Database communication have wrapper modules that you can download and use that are distribted through CPAN. Just like Java, it's a system dependant library that lets you call the functions. Perl has a utility (also called CPAN) for downloading, building and installing them. I just happen to be in the process of writing a Perl to C wrapper module so my mind jumped right to building your own.
First let me talk about this: > A language isn't fast, only it's compiler has had more time to be optimized. Natively > compiled programs in fact have less of a chance to be fast for these reasons:
Java advocates have been saying for years that Java can perform even better than C in certain instances. Maybe it can, but it doesn't! Every program I have ever seen that was written in Java was painfully slow. Load times suck, memory footprint is bad and screen interaction is slow. The best place I have seen for Java is running server side web based applications i.e. Tomcat. I personally prefer mod_perl since I'm a perl guy but Java is quite good in that space because it's bloated memory footprint and horrible load times aren't an issue.
To my main point:
There are more reasons than just performance to not use "modern" languages. You also have to consider code reuse. I'm a perl programmer. I can use any C libraries in perl by using one of the pre-built modules on CPAN or, if there isn't one already built, by writing my own XS interface. You can't use code written in java in a program written in perl, or for that matter, python or VisualBasic or C or C++. It's just not practical.
When you are looking at portability between "languages" vs operating systems, C is by far the most portable language.
People like to think of the language as a box, a complete package and like to say how incredibly good that language is. The problem is that no matter how big your language's box is, C's is bigger. Every language is, by nature, a subset of C since none can replace C and all can use things written in C. C is the least common denominator and so everything that needs to be language neutral needs to be written in C. Therefore general purpose libraries are written in C including security sensitive libraries.
I suggest the best way out of this would be a new language that was designed to be able to call C and be called as if it were C. It would need to compile to shared and static libraries that can be called directly from C or from other language's with their C interface. It would also need to be able to directly call functions in C libraries without interface routines.
Given those requirements, it's hard to imagine a language that looks or acts much different than C. I believe a language with a syntax similar to perl could do it. It would definitely have to lose `eval` and some other run-time stuff (require etc.) but a lot of the principals could carry over. That would be quite cool. C is a pain in the ass.
Re:And they don't want democracy so this will be b
on
Strike on Iraq
·
· Score: 1
Suicide bombing is wrong. I certainly don't contest that.
I believe that the buldozer driver didn't intend to kill the girl. That's not the issue.
Isreal does horrible things to the Arab people in Palistine. It's not OK to do monsterous things to a group of people because you think you are entitled to revenge.
In the case of suicide bombers, the person guilty of the crime is dead. It's human nature to want some sort of revenge or justice when something horrible like that happens but the fact is the person who did it blew themselves up. You should be asking yourself why they had so little hope that blowing themself up was the best option in their life and work on doing what you can to fix that.
Isreal is has engaged in ethnic cleansing. Isreal has put the Palistinians in "camps" and murderd them. Never again doesn't mean "never again unles we're the ones doing it". As long as we continue to fund these actions arabs in the middle east won't trust us.
Re:And they don't want democracy so this will be b
on
Strike on Iraq
·
· Score: 1
Right. That college girl they crushed with a bulldozer was probably in self defense. Give me a break! They operate with complete dissregard for human life and they need to be stopped.
And they don't want democracy so this will be bad
on
Strike on Iraq
·
· Score: 1
People in that part of the world think that the US is just out to make itself more wealthy at their expense. They see our support of Israel, a country that is doing some truely horiffic things, and they can't believe that we are trying to do what is good and right.
As a US citizen, I'm a lot more afraid for my safety now than I have ever been.
Even if everything goes perfectly during the war and post-war reconstruction (which it won't) the aftermath will be bad. The effects of this will be much larger than Iraq. We are doing something that the rest of the world asked us not to do. Never mind that it probably is the right thing to do, we were asked not to and we went ahead and did it because *we* wanted to. The US has shown the world that we don't care what they think, we'll can and will use our power however we like in order to stop people *we* think are evil. If I lived in another country I would be scared of the US and very pissed off.
I hope people realize that this is George W. Bush's bad choice, not all of America's. I'm embarrased that we elected him. The American people can no longer be trusted to only elect people who are qualified to be the "leader of the free world." By placing this horrible person in control of our military we've shown the rest of the world that they have to protect themselves from us. How will they do it?
I can't imagine people in the middle east would trust any government we setup or trust us for that matter unless we stop supporting Israel. Those bastards killed a girl with a bulldozer just this week, an American college student who was over there protesting. They repeatedly use their military against people who have no military of their own to defend themselves and seem to care little that they kill innocent people in the process. I know they got rounded up, tortured and killed in WW2 but that doesn't give them the right to do it to someone else. I believe it's this fundamental wrong which is fueling much of the anti-american sentiment in the middle east and we needed to end it before we did something stupid like invading Iraq.
With ReplayTV the hard disk module is the interface module and the digitizing module. You can watch anything recorded on one ReplayTV from another over 10/100 ethernet (They use HTTP over TCP/IP for delivery).
There are lots of flaws with the ReplayTV system but it sounds better than this tripe. 4 different modules all complete computers. Sounds way too expensive to manufacture to be competitively priced and it doesn't sound all that convenient either. Who funds these people?
Yea. That would make it hard to implement a "no Microsoft" policy. I agree with the underlying theme that getting rid of Microsoft software is a good idea. I don't think the hardline approach is necessary though. It will just happen on it's own using natural market forces.
Of course if a University was looking to pretty much get rid of all their Microsoft anyways they could use that money to bribe the software makers of those few titals that don't work on Linux to port them.:)
The post didn't say you have to rip out everything Microsoft and replace it with Linux. It says that they would have to agree to not buy anything from Microsoft. True, Linux is the obvious choice for new installations since it appears to be the natural successor to Windows but it could be MacOS or Sun machines or they could just keep nursing their current Microsoft software along for 10 years.
Licenses for existing software wouldn't have to be thrown out. All this does is change your buying decisions in the future. If you want to buy a new version of Windows only software you have to make sure you have Windows to run it on. If you want to run it on new hardware with some other OS, the software has to support the other OS. Eventually you would probably want to do a general roll out of a new standard OS but you could put it off as long as you would like. You wouldn't even have to do it at all if you don't want to. You'd just be stuck with very old Windows for a little while.
The support staff would simply have to learn how to admin RedHat 8 or MacOS instead of having to learn how to admin Longhorn. It's not that hard. RedHat is a breeze to install and day-to-day Linux admin is much less time consumig than Windows admin. Hiring one person who has in-depth knowledge of Linux would help steer things in the right direction and help you out of jams but for the most part you could probably get buy with the same people you have now.
I'd like to clear up some misconceptions that some Microsoft advocates have. First, Microsoft is a comodity software company. They rose to the top by making the standard software tools that everyone needs and charging less than everyone else. Second, Windows isn't popular because it is a great operating system. People forget quickly that when Windows expanded to dominate the computer industry it was horrible. The reasons it won against it's arguably superior competition were: 1) It could do pretty much everything that people needed to do 2) It did it cheaper than everyone else and 3) people could trust that they would get a good deal buy buying Microsoft stuff. None of those things are true for anymore. Linux and OpenOffice now win on all 3 points. 2 and 3 they win hands down. You can't beat free for a price and you can't beat the open source for trusting you won't get screwed. The only remaining issue is, is the software good enough? Does it do pretty much everything that people need it to? It's getting harder and harder to claim they don't. They now hold the powerful commodity software position. Microsoft is now suddenly in the premium software business and they don't know how to do that. Just like it took Microsoft a long time to displace the bloated behemoths of it's time it will take OpenSource software some time to displace the Microsoft behemoth but it will win. It's as sure a bet as Windows was 15 years ago.
No one expects you to throw out perfectly good Office 2000 suites on everyone's machines and install OpenOffice instead but when it comes time to put together a new 50 machine computer cluster are you really willing to dump $25,000 extra down the drain for pretty much the same functionality. We are living in an age where that's enough to buy the *entire* 50 machine cluster.
...just because you can see that a technology is going to hurt your business doesn't mean you should try and fight it. Throwing millions of dollers twords trying to make your customer's lives worse isn't going to help your business.
PVR's are going to kill the TV industry. We must stop them!
Linux is going to kill Microsoft. We must stop it!
File swapping is going to kill the music industry. Destroy it!
VCR's are going to kill the movie industry.
Video killed the radio star.
When did we go from a country where companies were supposed to compete on merrit to a country full of whining baby companies that don't want to change, inovate, or suffer any losses. Just because you can see that something is going to happen that will hurt your business, doesn't mean it's your job to try and stop it from happeneing. Lay a few people off to prepare for the impending belt tightening, don't hire a fleat of lawers and lobyists and wonder why your profits are down so much.
If you sell water don't sue mother nature expecting to stop the coming rain.
Slashdot Mirror
The problem is that the capacity to handle trafic on a highway decreases as people slow down. A highway that is near or at capacity will end up at a standstill in places if people slow down at any point.
If your highway is really that full, you need eather 2 more highways of that size or 3 times as may lanes on that highway. 8 each way sounds like a lot but it's no where neer as much as you should have. It should be 24 lanes each way or even better, there should be 2 other highways running parallel to it a few miles on eather side.
If you look at that A&M study carfully you will also notice that the green part of the graph, how much people drive, was completely unaffected by both congestion time and roadway expansion. People go where they need to go regardless of weather you build the road to handle the capacity well.
I've said this in other posts but the thing people don't realize is that when a traffic jam happens on a highway that road is handling many times the traffic it should. Safe travel distance on a highway according to most state agencies is 3 seconds. You should pass a fixed point on the ground 3 seconds after the back of the car in front of you. On a busy highway, there ends up bing more like 1 second or 0.5 seconds between cars. If you want that highway to be safe again you need to more than tripple it's size. Seriously. Adding a lane or two to a full highway is not enough. You end up with
I'll say it again:
The "more than" comes from the fact that you should have a mostly empty passing lane and an entrance/exit lane. Also, as is mentioned in the A&M study, when a road is at capacity people avoid it. They switch to 1) other times, 2) other roads and 3) other modes. When you fix the road, people switch back.
In 1999, the Surface Transportation Policy Project (STPP) analyzed data obtained from the Texas Transportation Institute and concluded
Texans don't know how to build roads
> There's gotta be new thinking in people moving...focusing not just on environment, but quality of life and practicality,
Absolutely. Let's not stick our heads in the sand for the next 10 years like we have for the last 10. The answer is to dramaticly increase the size and number of our highways.
We are *not* doomed to having undersized roads. Most highway expansions are done because there is congestion or traffic jams.
The problem is that by the time a highway gets to that point it is *way* over capacity and they end up expanding a tiny fraction of what is needed.
What most people don't realize is that when a highway gets to the point of having traffic jams just how far over capacity it really is. Heavily traveled highways should have a spare lane for passing on the left and a spare lane for entering and exiting on the right and middle lanes with enough room so the traffic on them can pass with 3 seconds between the back of one car and the front of the one behind it (This is the distance most states recommend for highway travel). That's about 3.5 seconds per car (from the front of one to the front of the one behind) and 4 seconds per truck.
Using this formula a highway that can carry 1 car per second should be 5-6 lanes. Most highways don't start bunching up and slowing down until they are over 1 car per second per lane (this is an educated guess based on observations of traffic in the Boston and New York area where people travel very close in rush hour). This means they are running at about 3 times what a "safe" highway would carry not counting the entrance and passing lanes. A 2 lane (per side) highway that carries 2 cars per second at peak should really be a 8 lane highway (16 lanes counting both sides). A 4 lane highway that is packed should really be a 14 lane highway (28 counting both sides) or more appropriately, 3 5 lane highways. Instead, projects usually add a paltry lane or two which just makes for a highway that is still over full it just has a little more room between the packed in cars which makes it slightly safer.
The problem with highway expansions is that the scale is way off. They don't try to get back to a safe comfortable traffic pattern, they just try to stop the traffic jams. This is a mistake. It makes for roads that are still dangerous, stressful and just end up needing to be expanded again a few years later.
If this sounds like a lot of expensive roads it's not. The expensive roads are the back roads. Highways are free. Most states pay for the building and maintainance of roads with gas tax. The highways with such high levels of traffic make much more in gas tax money than they cost. Most busy highways carry over one car per second. In contrast, those back roads that carry one car per minute or one car per hour still need to be built and tarred and plowed at great expense yet pull in very little gas tax money. Those are the expensive roads, not the highways.
As far as where to put new highways. Build them in a grid. North south highways intersecting with east west ones. Distance between them should depend on population densities. If there are buildings in the way, knock them down. Really. Knock them down. Don't build elevated highways around them, don't dig underground, take the land using eminent domain, knock the buildings down and put roads there. Then, take all the land next to the highways too and lease it back to people temporarily and kick them off when the buildings get old or it's convenient so you can expand the highway next time. In the end, it's a lot better economically to do that than making tens of thousands of people sit in cars with the engine running for an extra 20 minutes every day. Good efficient transportation makes for a clean happy prosperous city. It's worth it.
One of the interesting things about analyzing traffic using the cars/second/lane method is that you'll notice that as traffic slows down the capacity of the highway gets lower. That's because the time that it takes the car to go by g
> And what's gonna happen when you widen the roads? More people are
> going to move out further from the city, so more people are on the
> road. It's a self-defeating solution.
This is exactly the mentality I was talking about. The roads are too full so lets do nothing. It's not clever people, it's stupid.
Let's not stick our heads in the sand for the next 10 years like we have for the last 10. To understand why not building roads doesn't fix the capacity issue think about it the other way around. If you get a great job somewhere where you need to commute on a rather busy highway will you turn the job down? What if you have no other job opportunities? Don't kid yourself. Do you think the average person would take a job for $5,000 less per year to take 5 minutes off their commute? Our society is simply to specialized and spread out to not have good roads and it just isn't going to get less specialized or less spread out.
We are *not* doomed to having undersized roads. Most highway expansions are done because there is congestion or traffic jams.
The problem is that by the time a highway gets to that point it is *way* over capacity and they end up expanding a tiny fraction of what is needed.
What most people don't realize is that when a highway gets to the point of having traffic jams just how far over capacity it really is. Heavily traveled highways should have a spare lane for passing on the left and a spare lane for entering and exiting on the right and middle lanes with enough room so the traffic on them can pass with 3 seconds between the back of one car and the front of the one behind it (This is the distance most states recommend for highway travel). That's about 3.5 seconds per car (from the front of one to the front of the one behind) and 4 seconds per truck.
Using this formula a highway that can carry 1 car per second should be 5-6 lanes. Most highways don't start bunching up and slowing down until they are over 1 car per second per lane (this is an educated guess based on observations of traffic in the Boston and New York area where people travel very close in rush hour). This means they are running at about 3 times what a "safe" highway would carry not counting the entrance and passing lanes. A 2 lane (per side) highway that carries 2 cars per second at peak should really be a 8 lane highway (16 lanes counting both sides). A 4 lane highway that is packed should really be a 14 lane highway (28 counting both sides) or more appropriately, 3 5 lane highways. Instead, projects usually add a paltry lane or two which just makes for a highway that is still over full it just has a little more room between the packed in cars which makes it slightly safer.
The problem with highway expansions is that the scale is way off. They don't try to get back to a safe comfortable traffic pattern, they just try to stop the traffic jams. This is a mistake. It makes for roads that are still dangerous, stressful and just end up needing to be expanded again a few years later.
If this sounds like a lot of expensive roads think about trying to haul that many people using rails. A typical single track of railroad handles about 1/5th of the traffic of a single highway lane. Things scale much better if you replace an entire highway but even with optimistic estimates the cost would be astronomical. And once you have it all built you need to maintain all the stations, trains, drivers, track and it can't do the anywhere to anywhere that roads can do.
As far as where to put new highways. Build them in a grid. North south highways intersecting with east west ones. Distance between them should depend on population densities. If there are buildings in the way, knock them down. Really. Knock them down. Don't build elevated highways around them, don't dig underground, take the land using eminent domain, knock the buildings down and put roads there. Then, take all the land next to the highways too and lease it back to people temporarily and kick them off when the building
The space argument is silly. Most of the space taken up by roads is taken up by the little roads that go to everyone's door. How are you going to move things into/out of a house unless there is a road to it? Highways take up neglidgable space in comparison to how much traffic they carry. That back road to your door that carrys 1 car per hour, or 1 car per minute is much less efficient use of space and money than those highways that carry 3 cars per second. You want to save money and space, have local parking lots for every neighborhood and tripple the size of the highways. The only thing better than that would be to not have local parking garages, leving the local stuff alone and trippling the size of the highways.
> And what's gonna happen when you widen the roads? More people are
> going to move out further from the city, so more people are on the
> road. It's a self-defeating solution.
This is exactly the mentality I was talking about. The roads are too full so lets do nothing. It's not clever people, it's stupid.
Let's not stick our heads in the sand for the next 10 years like we have for the last 10. To understand why not building roads doesn't fix the capacity issue think about it the other way around. If you get a great job somewhere where you need to commute on a rather busy highway will you turn the job down? What if you have no other job opportunities? Don't kid yourself. Do you think the average person would take a job for $5,000 less per year to take 5 minutes off their commute? Our society is simply to specialized and spread out to not have good roads and it just isn't going to get less specialized or less spread out.
We are *not* doomed to having undersized roads. Most highway expansions are done because there is congestion or traffic jams.
The problem is that by the time a highway gets to that point it is *way* over capacity and they end up expanding a tiny fraction of what is needed.
What most people don't realize is that when a highway gets to the point of having traffic jams just how far over capacity it really is. Heavily traveled highways should have a spare lane for passing on the left and a spare lane for entering and exiting on the right and middle lanes with enough room so the traffic on them can pass with 3 seconds between the back of one car and the front of the one behind it (This is the distance most states recommend for highway travel). That's about 3.5 seconds per car (from the front of one to the front of the one behind) and 4 seconds per truck.
Using this formula a highway that can carry 1 car per second should be 5-6 lanes. Most highways don't start bunching up and slowing down until they are over 1 car per second per lane (this is an educated guess based on observations of traffic in the Boston and New York area where people travel very close in rush hour). This means they are running at about 3 times what a "safe" highway would carry not counting the entrance and passing lanes. A 2 lane (per side) highway that carries 2 cars per second at peak should really be a 8 lane highway (16 lanes counting both sides). A 4 lane highway that is packed should really be a 14 lane highway (28 counting both sides) or more appropriately, 3 5 lane highways. Instead, projects usually add a paltry lane or two which just makes for a highway that is still over full it just has a little more room between the packed in cars which makes it slightly safer.
The problem with highway expansions is that the scale is way off. They don't try to get back to a safe comfortable traffic pattern, they just try to stop the traffic jams. This is a mistake. It makes for roads that are still dangerous, stressful and just end up needing to be expanded again a few years later.
If this sounds like a lot of expensive roads think about trying to haul that many people using rails. A typical single track of railroad handles about 1/5th of the traffic of a single highway lane. Things scale much better if you replace an entire highway but even with optimistic estimates the cost would be astronomical. And once you have it all built you need to maintain all the stations, trains, drivers, track and it can't do the anywhere to anywhere that roads can do.
As far as where to put new highways. Build them in a grid. North south highways intersecting with east west ones. Distance between them should depend on population densities. If there are buildings in the way, knock them down. Really. Knock them down. Don't build elevated highways around them, don't dig underground, take the land using eminent domain, knock the buildings down and put roads there. Then, take all the land next to the highways too and lease it back to people temporarily and
Face it. Most people live in cramped spaces because they have to. All things being equal, they would rather have a yard. Would someone in Manhatten like having a big house and a big back yard to play in. Sure. It just doesn't make sense concidering the land would cost about 40 million.
> The car system is falling apart
The car system works well where it is done right. The problem we are having has nothing to do with highways not scaling well. They scale excellently, much better than rail.
The problem is that as a society, people are more specialized in their jobs and most households have two incomes now. The increased specialization makes commutes longer because we need to travel farther to reach a company that does what we do. Historicly we would just move to where the company is but since there is another specialized income in the household we just have to commute. If one person works an hour away from the other there is going to be at least an hour of commuting between the two. There is no way arround that.
This makes for more cars traveling further every day and traffic increases. This isn't going away. There is no magic bullet here to make roads unneccesary. We just need to make more roads.
Car's aren't perfect but they are the most economically efficient solution for most places. The main problem with cars is that most governments have decided not to improve the roads when improvements are needed. When they do improve them they do stupid things like the Big Dig in Boston. Trains are wildly expensive for anything but the most densely populated cities. Segway's are too slow to handle long distance travel. Cars are versitile, quick, efficient, and do their job well.
The problem with getting rid of cars is that I want a back yard. The bigger the better. Most people don't want to live on top of one another in big buildings with no place for their kids to play. A world without cars is a world where everyone needs to be packed in on top of each other so that mass transit can work. I don't like that idea.
If the roads are too crouded, build bigger roads. It's not a hard conept. Why do people think they're doing something clever by not building roads when they should (I live in New Hampshire, north of Boston where commuting is horrible.) We waste thousands of man-hours of time every day, waste tons of gas, increase pollution and make thousands of peoples lives more stressful. It's not celever!
Let's kick the smart people out after they've helped us out and built relationships with people here so businesses will be forced to higher people who have lived here longer. Absolutely stupid!
We should be doing everything in our power to get smart people to move to the US. It's what has made us strong in the past.
This article was written to try to get people to buy technology training. In reality I expect most companies to be reducing the staff of their IT departments to try to become profitable again.
Computer systems these days are built with more redundant and more powerful components that are easier to maintain. Most companies have switched away from the Windows 9x line which saves a awful lot of time dealing with stupid desktop issues. We have also seen the change to web-based software which is getting cheaper and easier to implement and support every day. It is now possible for a small group (3 people or so) to manage a large pool of inexpensive web-servers (20-50) which supports a huge application used by thousands of people. This model is increasingly being used and it works well and saves bundles of money. This, combined with companies new-found zeal for cost cutting will drive the numbers of IT professionals down in the coming years.
My advice is if you aren't in IT, don't try to get into it unless you are really good with computers. If you are in IT but aren't very good at it, think about finding another career. The future in IT may not be very bright.
Today's computers have almost nothing in them that was available 2 years ago no less twenty. The core of a computer is the north bridge chipset. This is where most of the speed is determined and most of the cost comes from. This is where we have DDR Ram, 533 MHz Front side busses, and AGP 8x. Nothing here remotely resembles a PC from 20 years ago. Sure, computers still have a version of the keyboard port they used 20 years ago. We still use it because it's really good at being a keyboard port.
The PCI section was funny. In one breath the article said that PCI express is an evolution from PCI that is invisible to software. The quote was: "mainly a hardware change that will result in simpler motherboard and peripheral designs". Then 5 lines down the article said that when PCI Express is adopted "a whole new class of PC will emerge." Yea, and that class will be slightly different than the class before just like always.
As far as the claims that the hard drive attachment technology hasn't changed much in the last 20 years it's very hard to find anything in modern IDE that existed back in the PC. The physical signaling is very different, the controller is on the drive now, there is a protocol (ATA) running on top of the bus, the addressing has completely changed. Iâ(TM)d say the biggest change with IDE came back around 1993 when ATA was developed to run on top of it. I am a great fan of SerialATA but it is just an evolutionary change in the physical communications layer. That's one of the best things about it, that it is compatible with the "legacy" architecture and yet the article raves about it and then laughably backes it up by saying that the first serial ATA drive out was "quieter and cooler-running than its classic ATA counterparts" Pure fluff.
As for the floppy, it is certainly time for something to be done about it and yet next to no work has been done on a replacement. The floppy disk is a random read-write bootable removable medium that every PC operating system natively supports. There is no other device that can claim that. CD Burners should have replaced the floppy years ago but the manufacturers never got together and built a new standardized low-level interface. Even bootable CD's still emulate a floppy disk and the boot image is limited to the size of a 2.88 MB floppy. The floppy replacement is an issue that now *needs* to be addressed and yet the articleâ(TM)s suggestion is to simply leave it out without anything to replace it's unique functionality.
Every once in a while these fluff articles pop up. "Soon computers will be as simple, cheap and as easy to use as your phone" they spout "and all they need to do is leave out all that old stuff that you don't really need". The thing they seem to miss is that it has already happened. You can go down to the store and buy a nice legacy-free computer with none of those useless 20 year old keyboard ports or dumb serial ports and it's cheap and easy to use and it's a palm pilot and it sucks for doing what computers are good at. There are all kinds of "legacy-free" computers out there, Ipaq, Tivo, smart phones, there's even those super-cool 3com Audreyâ(TM)s that are all the rage because they are legacy-free
Legacy free usually means not compatible with the old stuff and for a computer that means it's less flexible and thus less powerfull and less desirable. There is a *huge* amount of effort that has gone into designing and supporting these "legacy" systems and to suggest that because it's old it should go is to forget a fundamental truth in the PC industry:
> At a conservative estimate, MySQL is 25 years behind the state of the art.
But state of the art isn't what's important with OSS. OSS is about the commodity market and relational databases *are* a commodity now. True, not all the features you need are in both of these databases and it's easy to come up with a feature list where Oracle looks great. That's not the point. When you need a database for a project odds are one of these two (PostgreSQL or MySQL) will give you what you need. A lot of programmers don't like the idea of learning how to code against these two because they already know Oracle and with that knowledge they can tackle any project. The problem is the guy next to you who knows these two OSS databases can tackle that project for $10K less. Who's more valuable as a programmer? The answer isn't always going to be the OSS programmer but it will more often than not.
These databases are especially important for commercial applications that need a DB back end. It's one thing to have a $2-$10K db license that you can share among multiple applications but when you are selling a product that relies on a database backend, using an OSS DB can save you the cost of the comercial database hundreds or thousands of times over.
Ellison will still make money for years to come just like Apple has and Microsoft will. Being relegated to the non-mainstream doesn't mean death, it just means there is a new set of rules. No more rolling in wheelbarrows full of cash by charging big money for commodity software.
Wow! 3DFx was an impressive company with really good technology. I was a NVidia man my self but up until the end I debated the wisdom of my choice. 3DFx had sound technology, loyal customers, and some really powerful IP (their multi-chip rendering was by far the best.) It takes a real moron to run that strong of a company into the ground. It would seem from the outside that they could have backed off, sured things up and lived to see another day. I can't believe someone else hired him! What the hell are these people thinking?
I chose ReplayTV over Tivo a long time ago. I'm a big Linux fan but I just couldn't get over the 10/100 ethernet port and all the cool options it represented. They used http for interaction with the box which is wonderfully flexible. It was a nice modular architecture with lots of expantion possibilities. And months went by and nothing happened. And more months and more nothing. I haven't had a software update on the box for almost a year and there are issues that need to be fixed. The company has been complietely impotent.
In my experience, there is a type of person who loves to be in charge and should never be allowed to do so because they suck at it. They see most successes as their's and most failures as someone else's. To them, every prolem has an easy solution and it's the best one. I have been blessed to work with some great people who know how to encourage, lead by example, trust in the wisdom of the people who know what they are talking about, and stick to what they do best.
You can blame these things on bad engineers, small budgets and other logistical issues but let's face it: A good leader *makes* things happen. An average leader *let's* things happen. It takes a bad leader to stop things from happening and I get the feeling that's the type of leader Greg Ballard is.
> Anybody but Bush in 2004
Oh my god! I want this bumper sticker. Does anyone know who can print up a nice red white and blue version of that?
Of course Java can call native libraries, that was part of my argument too. I don't hate Java, it's just not appropriate for libraries that you want to share between languages. C is the only option there which is why SSL support, rsync support, DB interfaces, OpenGL etc are all written in C. Things written in C can be used in any language. You can't say that about any other language.
A good example of this is ImageMagick. It's a nice library for manipulating images. It's written in C but you can use it from Perl, Python, PHP, C, C++ and Java. If it was written in any other language it would only be usable in that language.
C is prone to memory issues because the standard way it deals with strings is horribly dangerous. Strings shouldn't be represented by a simple pointer to a place in memory. Strings should, at the very least, be structurs with information on length of buffer space and length of the currently stored string. All functions that operate on strings should gracefully handle trying to read or write past the end. Many libraries have already implemented passing a length in with a string when one is passed in but this is not done in a standardized way. We need to get rid of null terminated strings in C.
And putting on my Perl advocate's hat, I doubt it's easier to invoke C from Java than Perl. I've heard that Perl has the tightest C integration of any modern portable language. Many simple C functions are directly accessible from perl (printf, sprintf, select). Most are available by using one of the standard modules included in the Perl 5 distribution. More advanced things like SSL, Rsync, ImageMagik, and Database communication have wrapper modules that you can download and use that are distribted through CPAN. Just like Java, it's a system dependant library that lets you call the functions. Perl has a utility (also called CPAN) for downloading, building and installing them. I just happen to be in the process of writing a Perl to C wrapper module so my mind jumped right to building your own.
First let me talk about this:
> A language isn't fast, only it's compiler has had more time to be optimized. Natively
> compiled programs in fact have less of a chance to be fast for these reasons:
Java advocates have been saying for years that Java can perform even better than C in certain instances. Maybe it can, but it doesn't! Every program I have ever seen that was written in Java was painfully slow. Load times suck, memory footprint is bad and screen interaction is slow. The best place I have seen for Java is running server side web based applications i.e. Tomcat. I personally prefer mod_perl since I'm a perl guy but Java is quite good in that space because it's bloated memory footprint and horrible load times aren't an issue.
To my main point:
There are more reasons than just performance to not use "modern" languages. You also have to consider code reuse. I'm a perl programmer. I can use any C libraries in perl by using one of the pre-built modules on CPAN or, if there isn't one already built, by writing my own XS interface. You can't use code written in java in a program written in perl, or for that matter, python or VisualBasic or C or C++. It's just not practical.
When you are looking at portability between "languages" vs operating systems, C is by far the most portable language.
People like to think of the language as a box, a complete package and like to say how incredibly good that language is. The problem is that no matter how big your language's box is, C's is bigger. Every language is, by nature, a subset of C since none can replace C and all can use things written in C. C is the least common denominator and so everything that needs to be language neutral needs to be written in C. Therefore general purpose libraries are written in C including security sensitive libraries.
I suggest the best way out of this would be a new language that was designed to be able to call C and be called as if it were C. It would need to compile to shared and static libraries that can be called directly from C or from other language's with their C interface. It would also need to be able to directly call functions in C libraries without interface routines.
Given those requirements, it's hard to imagine a language that looks or acts much different than C. I believe a language with a syntax similar to perl could do it. It would definitely have to lose `eval` and some other run-time stuff (require etc.) but a lot of the principals could carry over. That would be quite cool. C is a pain in the ass.
Suicide bombing is wrong. I certainly don't contest that.
I believe that the buldozer driver didn't intend to kill the girl. That's not the issue.
Isreal does horrible things to the Arab people in Palistine. It's not OK to do monsterous things to a group of people because you think you are entitled to revenge.
In the case of suicide bombers, the person guilty of the crime is dead. It's human nature to want some sort of revenge or justice when something horrible like that happens but the fact is the person who did it blew themselves up. You should be asking yourself why they had so little hope that blowing themself up was the best option in their life and work on doing what you can to fix that.
Isreal is has engaged in ethnic cleansing. Isreal has put the Palistinians in "camps" and murderd them. Never again doesn't mean "never again unles we're the ones doing it". As long as we continue to fund these actions arabs in the middle east won't trust us.
Right. That college girl they crushed with a bulldozer was probably in self defense. Give me a break! They operate with complete dissregard for human life and they need to be stopped.
People in that part of the world think that the US is just out to make itself more wealthy at their expense. They see our support of Israel, a country that is doing some truely horiffic things, and they can't believe that we are trying to do what is good and right.
As a US citizen, I'm a lot more afraid for my safety now than I have ever been.
Even if everything goes perfectly during the war and post-war reconstruction (which it won't) the aftermath will be bad. The effects of this will be much larger than Iraq. We are doing something that the rest of the world asked us not to do. Never mind that it probably is the right thing to do, we were asked not to and we went ahead and did it because *we* wanted to. The US has shown the world that we don't care what they think, we'll can and will use our power however we like in order to stop people *we* think are evil. If I lived in another country I would be scared of the US and very pissed off.
I hope people realize that this is George W. Bush's bad choice, not all of America's. I'm embarrased that we elected him. The American people can no longer be trusted to only elect people who are qualified to be the "leader of the free world." By placing this horrible person in control of our military we've shown the rest of the world that they have to protect themselves from us. How will they do it?
I can't imagine people in the middle east would trust any government we setup or trust us for that matter unless we stop supporting Israel. Those bastards killed a girl with a bulldozer just this week, an American college student who was over there protesting. They repeatedly use their military against people who have no military of their own to defend themselves and seem to care little that they kill innocent people in the process. I know they got rounded up, tortured and killed in WW2 but that doesn't give them the right to do it to someone else. I believe it's this fundamental wrong which is fueling much of the anti-american sentiment in the middle east and we needed to end it before we did something stupid like invading Iraq.
With ReplayTV the hard disk module is the interface module and the digitizing module. You can watch anything recorded on one ReplayTV from another over 10/100 ethernet (They use HTTP over TCP/IP for delivery).
There are lots of flaws with the ReplayTV system but it sounds better than this tripe. 4 different modules all complete computers. Sounds way too expensive to manufacture to be competitively priced and it doesn't sound all that convenient either. Who funds these people?
Yea. That would make it hard to implement a "no Microsoft" policy. I agree with the underlying theme that getting rid of Microsoft software is a good idea. I don't think the hardline approach is necessary though. It will just happen on it's own using natural market forces.
:)
Of course if a University was looking to pretty much get rid of all their Microsoft anyways they could use that money to bribe the software makers of those few titals that don't work on Linux to port them.
The post didn't say you have to rip out everything Microsoft and replace it with Linux. It says that they would have to agree to not buy anything from Microsoft. True, Linux is the obvious choice for new installations since it appears to be the natural successor to Windows but it could be MacOS or Sun machines or they could just keep nursing their current Microsoft software along for 10 years.
Licenses for existing software wouldn't have to be thrown out. All this does is change your buying decisions in the future. If you want to buy a new version of Windows only software you have to make sure you have Windows to run it on. If you want to run it on new hardware with some other OS, the software has to support the other OS. Eventually you would probably want to do a general roll out of a new standard OS but you could put it off as long as you would like. You wouldn't even have to do it at all if you don't want to. You'd just be stuck with very old Windows for a little while.
The support staff would simply have to learn how to admin RedHat 8 or MacOS instead of having to learn how to admin Longhorn. It's not that hard. RedHat is a breeze to install and day-to-day Linux admin is much less time consumig than Windows admin. Hiring one person who has in-depth knowledge of Linux would help steer things in the right direction and help you out of jams but for the most part you could probably get buy with the same people you have now.
I'd like to clear up some misconceptions that some Microsoft advocates have. First, Microsoft is a comodity software company. They rose to the top by making the standard software tools that everyone needs and charging less than everyone else. Second, Windows isn't popular because it is a great operating system. People forget quickly that when Windows expanded to dominate the computer industry it was horrible. The reasons it won against it's arguably superior competition were: 1) It could do pretty much everything that people needed to do 2) It did it cheaper than everyone else and 3) people could trust that they would get a good deal buy buying Microsoft stuff. None of those things are true for anymore. Linux and OpenOffice now win on all 3 points. 2 and 3 they win hands down. You can't beat free for a price and you can't beat the open source for trusting you won't get screwed. The only remaining issue is, is the software good enough? Does it do pretty much everything that people need it to? It's getting harder and harder to claim they don't. They now hold the powerful commodity software position. Microsoft is now suddenly in the premium software business and they don't know how to do that. Just like it took Microsoft a long time to displace the bloated behemoths of it's time it will take OpenSource software some time to displace the Microsoft behemoth but it will win. It's as sure a bet as Windows was 15 years ago.
No one expects you to throw out perfectly good Office 2000 suites on everyone's machines and install OpenOffice instead but when it comes time to put together a new 50 machine computer cluster are you really willing to dump $25,000 extra down the drain for pretty much the same functionality. We are living in an age where that's enough to buy the *entire* 50 machine cluster.
...just because you can see that a technology is going to hurt your business doesn't mean you should try and fight it. Throwing millions of dollers twords trying to make your customer's lives worse isn't going to help your business.
PVR's are going to kill the TV industry. We must stop them!
Linux is going to kill Microsoft. We must stop it!
File swapping is going to kill the music industry. Destroy it!
VCR's are going to kill the movie industry.
Video killed the radio star.
When did we go from a country where companies were supposed to compete on merrit to a country full of whining baby companies that don't want to change, inovate, or suffer any losses. Just because you can see that something is going to happen that will hurt your business, doesn't mean it's your job to try and stop it from happeneing. Lay a few people off to prepare for the impending belt tightening, don't hire a fleat of lawers and lobyists and wonder why your profits are down so much.
If you sell water don't sue mother nature expecting to stop the coming rain.
Making a kickstart install is a brease and there are no arbitrary hoops to jump through.
Being a sysadmin is hard enough. We don't need MS intentionally making it harder.
When you can buy a computer for $150 that does everything you need it to, what will Microsoft do to convince you to give them money?