I'm one of the people who runs Windows 2008 Server on a laptop, and I can tell you now, it's Vista with a higher retail price-tag.
Microsoft has been releasing server builds that are virtually identical to the desktop editions for years now. Windows XP 64-bit uses Windows 2003 service packs. Windows 2008 uses Vista drivers, the server editions have the DirectX gaming APIs, the workstation editions can serve file and web pages, etc...
The only difference in server is different initial default settings, some additional modules (such as Active Directory), and a half-dozen tuning parameters, most which are set through the registry anyway.
Even Active Directory will run on a desktop OS! It's called Active Directory Application Mode (ADAM).
...construct a solution tailored to a specific environment...
The thing is, I'm a systems integrator, and as a part of my consulting job I have worked on small and large projects at something like 100 different companies and organisations of all sorts. Do you know what I discovered?
Companies all have the same needs when it comes to IT.
They all have the same need for home directories, application deployment, patching, security, backup, and the like. They all run the same kind of applications for the same reasons. It doesn't change as much as people think from company to company. So why should I have to reinvent the wheel every time? Why shouldn't I use a pre-packaged solution which does exactly what everyone needs? Isn't that a huge waste of time and money?
Notice that the Munich project ran over budget.
This is why Microsoft is so hugely successful. They realised that people don't want to "do it themselves" unless it's a hobby.
Re:The grand-daddy of them all...
on
Web Singletons?
·
· Score: 1
I think the thing with teaching is to look at different approaches taken by other people, and use the method that actually produces results. I saw teachers try several different techniques at high school and university level, and they all failed miserably, except for one guy who was wildly successful. He was actually tasked with teaching us Haskell, which is an abstract pure functional language that most people "don't get". Despite that, he successfully taught it to the majority of students in his class, who then couldn't learn Basic or C from other teachers in the subsequent semesters.
His technique was simple. He set up a digital projector with his laptop so he could sit and face the class, and projected the development environment onto the screen. He then started coding really simple problems ("add all the numbers in a list") and running the example code, often through the debugger.
With almost every problem, he'd deliberately make a mistake (leave out a bracket, or not increment the counter in a loop, etc...) and demonstrate how the program failed, how to interpret the error messages or debug it to find the source of the error, and the result of fixing the error. He'd involve the class, asking the students to 'spot the error'.
I could already program when I started his class, so it was fascinating to watch him teach. What I found is that he was teaching exactly the things that I found difficult when I first started out:
- It's hard for beginners to go from a compiler error message to a fix. I still remember the Borland C++ error message "Cannot assign value to rvalue". It took me a day to figure out that I had an "=" instead of an "==" in an expression.
- Writing programs is relatively easy to learn, but learning how to "mentally execute" a program step-by-step to debug it, or use a debugger, is hard. Many beginner programmers just assume that it'll do what they meant, not what they actually wrote. Writing correct programs is much harder.
When I was studying computer science at university, I had read about all these fancy cryptographic techniques, and I imagined that banks were these encrypted, firewalled fortresses of IT security, monitored by the most competent, most vigilant administrators.
I was very wrong.
Let me tell you about my experience of IT security in banks.
A couple of years ago, I was sent to one of Australians largest banks. I was there for a 1 week engagement to install the latest virus scanner software on some servers. It sounded like a great opportunity to have a look at some high-end systems and see how they were managed. So I turn up in the morning, and start unpacking my laptop, when the project manager warns me:
"Don't plug your laptop into the network. We have to make sure you have the latest patches and AV first."
I fully understood his position, of course, they couldn't just let some random guy plug some a laptop into the network. It was a bank after all, security matters. I was Wrong. He corrected me:
"Oh no.. that's for your own protection! There's hundreds of viruses on this network, if you plug an unpatched machine into it, it'll be infected in seconds."
I was stunned. He wasn't even joking. I did plug my laptop in (which was well patched), and ran Ethereal for a few minutes, during which time I saw several viruses attempting to hack my machine. It was incredible. I've never seen that kind of attack rate anywhere, and I've been to large, unfirewalled university networks and school networks.
In fact, I didn't even really need to plug myself in. There was a WiFi connection available, with an easily recognisable SSID (the name of the bank). Of course, it was unencrypted, unsecured, and plugged directly into the desktop LAN.
Next, I got a tour of the data center, which was an eye-popping experience in itself. The bank had recently invested in fancy new retina-scanning door locks. It looked like it was straight out of a James Bond movie. However, it was taking too long to program in every person who needed access into the system, so they had simply propped the door open with a bucket. The inside of the room was just as scary. I walked past DOS machines, Windows 95 "servers", and I saw at least one NT 3.x machine. This was in 2005.
Eventually, I got around to planning the AV software upgrade. Except it wasn't. It was first-time-install, because the majority of their servers had no AV. The amount of work required to verify compatibility during for a the rollout was deemed too expensive, and I never did get to install the AV software. They did buy the licenses though, so it's entirely possible they installed it themselves. It's possible, but I wouldn't bet on it.
As a preface, "How much physics do you know?": A lot.
Anyway, there was a paper in Astrophysical Journal a few years back, in either 2005 or 2006 from the University of Victoria. They got nice rotation curves for galaxies just from general relativity, without invoking dark matter. Kind of neat.
Mike.
I know a bit of physics too, having studied it at University, and Dark Matter always gave me the impression that when Physicists computed Galactic rotation, they did the typically lazy "good enough for Physics" mathematics and simplified -- they used Newtonian mechanics. I have not seen even one paper or article mentioning relativity. Not surprisingly, at the galactic scale, Newtonian mechanics doesn't quite work out. Soo.. wait for it.. what's the correct answer? There must be a mysterious, magical force out there! It can't possibly be that we used the equations that are generally known to be inaccurate. No sir. No way. This, err.. inaccuracy is caused by something I just discovered! Can I have my Nobel now?
30 TB per night sounds like a lot, but 1.5 TB drives are about AUD 350 each, retail. By 2016, I'd expect vendors to have released at least a 10 TB hard drive at that price point, and I wouldn't be surprised if we're using 30 to 50 TB drives.
So it all boils down to about $1000 per night of operation, or about $350K per year. Not exactly expensive for a science project. A single mars mission costs about $300M, but this telescope would generate more discoveries. That's not even considering that storage costs would continue to drop over the lifetime of the telescope, so the eventual total cost may be less than $100K per year. That's the salary of just one person!
Stop quoting nonsense you heard from your grandpa.
Film is a terrible archival medium, except for maybe silver based black and white film. It fades, the color changes, is easily damaged, and the original degrades when copied. George Lucas has spent $millions carefully restoring the archived Star Wars films, and they're a lot less than 50-60 years old. Film over 50 years old usually takes heavy processing to be even watchable.
On the other hand, digital archives are trivial to copy losslessly, so there's no need for any physical media to last for the length of the archival time.
Unfortunately the parent post is right - space colonization in the foreseeable future is unlikely for many reasons that somehow seem glossed over by the "true believers".
Lets face the cold hard reality of space - it is both cold and hard. There's nothing out there but rocks. Nobody wants to live on cold hard rocks. Some people might go there for science, or out of the curiosity of a tourist, but nobody will ever want to make a life there.
I'm sure of this because people already have the opportunity to go live on cold hard rocks that are far from civilization, right now, but don't. There are huge, unclaimed tracts of rocks right here on Earth. Most of Canada, Russia, and Antarctica is an uninhabited wasteland, despite massive overpopulation in other parts of the world. Heck, most of the continent of Australia is an uninhabited desert, so there's also a choice of warm hard rocks, if that's your preference.
Compared to any other place in space, these places aren't even that uninhabitable. There's usually some water, a breathable atmosphere, soil, gravity, real time telecommunications, and resources and equipment can be brought in a cost of mere thousands of dollars instead of billions. Yet despite these manifold advantages over space colonization, there's no popular demand for massive government spending to colonize these places. Why not?
Sure, it's not glamorous, but we can do it right now! We could, if we wanted to, colonize Antarctica. It wouldn't even be that hard, all the technology is available right now. We could move a billion people there if we had to. Does anybody want to though? Do you? Would you, right now, give up where ever you are, with your job, friends, family, and go live in a place like that? Or if you don't like the cold, you're welcome here in Australia! The desert has some really cheap land. You can buy a place the size of a small American state if you want to. Oddly enough, most of tens of thousands of immigrants that come to this country every year go to live in the larger cities. Not a huge demand for desert living for some strange reason.
Now let me put it this way - Mars is just like Antarctica, but much colder, much more remote, much harder to reach, much harder to come back from and there's no atmosphere. Not to mention that the return ticket for a family sightseeing tour these days is $1 Trillion*.
Living on Mars is totally irrational wishful thinking. Unless some miracle occurs like the sudden invention of cheap wormhole generators or anti-gravity or some similarly quick and easy way to get about becomes available, I just don't see it, and we can't make plans where step #1 depends on a miracle. Even given some cheap magic space transport device, every destination is still "Not Earth". Step #2 will be that anywhere we might want to go is instant-death-to-the-unprotected. Every little detail is lethal. Did you know that due to the lack of weathering, the dust on the Moon is microscopically jagged and razor sharp? It'll cut your lungs up if you breathe it in for any length of time. Just look up "silicosis" to find out all about the joys of that particular ailment. In comparison, the Australian Outback is so inviting that even the dust comes in a lung-friendly rounded format, but there's still places you can go where the next nearest person is 100km away. I don't see that changing any time soon, so why would anyone expect space to be colonized first?
* An estimate only, one that NASA keeps revising upwards. Before betting on cheap space travel, wait for the prices to actually start dropping first.
While it is true that Windows Server 2008 is almost exactly the same as Vista SP1, down to the hotfixes and drivers, the tangible difference is really a bunch of compiler macros and flags that Microsoft charges hundreds of dollars for.
I run Server 2008 with the "Desktop Experience" pack as a substitute for Vista on my work laptop because of a bad experience I once had while doing a demo for a customer on an XP laptop - I had developed a simple ASP.NET website and was making a demonstration when one of the users had managed to produce a "HTTP/500" error. It was incredibly embarrassing to have my supposedly "highly reliable" system lock up after just a few clicks. It took me days to figure out that the "crash" was caused by a completely artificial limitation introduced by Microsoft into XP to differentiate it from their Server line - one of the TCP/IP connection limits was the culprit. I had never noticed it while developing, because loopback connections are not affected.
So now I run an MSDN licensed Windows 2008 as a "workstation" OS so that I can avoid the Microsoft Marketing Department's deliberately introduced bugs, leaving only the plain old technical bugs, of which there are thankfully fewer than some previous MS operating system releases.
If Microsoft made cars, the replies to a question such as this would inevitably be "I want a car that goes 300 miles per hour, 5000 miles to the galon, has room for all the family, and.. and looks awesome! Yeah! With bells..and whistles!".
Back to reality. What I would like from Microsoft is much simpler - I want them to do what they've delivered in the past but failed to do so recently. Microsoft's strength has always been consistency and compatibility. For example, virtually every platform called "Windows" has support for both Batch and VBScript automation. This is great, if as an admin, I want to write a script, I can. No problem, it will just work.
Recently they've introduced Powershell, which has made even Linux zealots a little jealous. Except it's not installed by default. That's right, in a 10GB base install of Windows Server 2008, they couldn't spare the 2MB scripting component. It couldn't make it in there. No room... or something. This means that unless a server requires Powershell (eg: because it runs Exchange 2007), it probably won't have it! This makes Powershell an order of magnitude less useful than it should be. Of course, you can install it with another script (a batch file, for example!), as long as you shove the DVD into the server.
So... you're a contractor, you're at a new site you haven't seen before but you know they have Windows 2008 on all their servers and you've been asked to bulk-reconfigure a bunch of servers on a deadline. You plan to use your Jedi PowerShell skills and your psexec bulk automation wizardry, don't you? Try again. It's back to Batch files for you, like everyone else.
This is why end-user ideas like "componentized operating systems" are stupid. Unless you're developing for the embedded market, shaving 1MB off a 10GB base install is just frustrating admins and software developers for no tangible gain. If a feature is not guaranteed to be installed, developers have to assume it isn't and hence can't depend on it or use it. Most users don't even know where their install DVD is, if they even have one, so they can't fix any missing components required by a new piece of software.
And Microsoft really has to drop the ridiculous "security through signature" crap. You have to sign Powershell scripts now. Think for a minute how ridiculous that is. You can run unsigned executables, binary blobs with any possible content at all, but not a script? How does that makes sense? The first thing every admin does is turn that feature off, resulting in a hassle and no additional security at all. Scripts, like everything else, are already restricted to whatever the user can do, they need no additional security. Not to mention that no-one would be stupid enough to run an script from an untrusted third-party without glancing through it first!
I just tried this with Mozilla 3.0 beta (Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.9b4pre) Gecko/2008021607 Minefield/3.0b4pre) and the score was 60/100. Not bad.
However, it doesn't look like the reference rendering. There's no colors, there are red blocks in a few places, and there's a purple "X" as well.
I hate this 'rule of thumb' that people keep trotting out that we only use x% of software, for some low value of 'x'. That's simply not true, so stop bringing it up like it's a valid argument for anything.
I keep hearing this bullshit if from Unix zealots, from people flogging 'thin' or 'web 2.0' products, and from Luddites that are 'perfectly happy' running WordPerfect 5.1 on their OS/2 machine.
Lets think of a simple scenario. Imagine a fictional company MiniSoft Software that makes a word processor. They advertise that their program has 100 features! Of course, you know that most users will only use about 10% of that most of the time, and maybe an occasional 1% rarely. So why have the other 89 features in there? Most users won't be using it!
What this kind of oversimplified 'analysis' misses is that that '1%' extra is different for every user. Glenda in marketing might use the 'mail merge' feature once a month. The payroll officer might have to use the database integration feature. The warehouse manager might be using the barcode printing. The international sales office might use the Unicode multi-lingual features.
Once you add up all of those '1%' pieces, all too often, you end up with... 100 features or so. This is why MiniSoft Office is so 'bloated'. Because somewhere, out there, there's someone who uses the macro functionality, or the right-to-left text input, or the dynamic forms, or... something. It's not bloat... it's what users expect from their software -- that the same consistent product be useful for all of the staff in an entire business.
So to reiterate, just because YOU only personally use the "bold" and "italic" buttons on the toolbar doesn't mean that someone else can get by with only those two buttons.
Get used to it, because software is only going to get bigger and more 'bloated', not less.
I keep hearing these excuses over and over, but they still make no sense.
There are no parachutes on airliners for the following reasons:
1. Parachutes are heavy, so a plane equipped with them could carry less cargo or passengers and ticket prices would go up. So? They already have life jackets on planes. Parachutes are not made of lead. the last time I put one on, it wasn't more than a few kg.
2. Parachutes are very complex to pack, and would have to be unpacked, inspected, and repacked at regular maintenance intervals, at considerable expense (not to mention increased time out of service for the plane). How does that make any sense? The plane doesn't have to be grounded to repack a parachute. They can be replaced with a set of already checked and repacked parachutes, and they can be repacked at any time while the plane continues on its merry way. They're called 'spares', heard of them?
3. If the plane is high enough that parachutes will be of any use, it's impossible to open most exit doors as pressure seals them against the inside of the fuselage. That's a design feature that can be changed. They're not designed to open because no existing airline uses parachutes. If airlines started using parachutes, the doors could be redesigned to open!
4. Only a tiny fraction of passengers would understand how to use parachutes. When all the others slam into the ground at terminal velocity -- especially if the plane somehow survives -- it's a brave new world of stupendously huge liability for the airline.
There is no user intervention required: They could be opened by a cord attached to the exit, just like the airforce uses for large paratrooper jumps.
And are dead passengers much better for liability? This is exactly like the car manufacturers dragging their feet on installing seatbelts. There's no excuse. If the parachutes could save lives in some circumstances, they should be installed. Life jackets on the other hand are useless. Planes hitting the water 'at speed' invariably disintegrate and kill all aboard. However, many 100% fatal accidents have been over land, with enough available time to get the passangers out. Parachutes could have saved hundreds of lives.
Think of it this way: In world war two, tens of thousands of allied soldiers were thrown out of planes, in the dark, while getting shot at. Most of them survived. Some people broke bones, some didn't have their chute open, but by and large, most people survived.
The hardest part to having people jump from a jumbo is getting 300+ people to put on their parachutes in a reasonably short period of time. That's the only thing I can think of that would be a stumbling block. Everything else can be solved.
I wish more artists would embrace realistic fantasy.
It sounds like an oxymoron, but it's the difference between a movie only kids could enjoy, and something adults would want to go see too.
The first thing that struck me about those pictures is that nobody would ever, ever, ever use one of those contrived contraptions in a battle. A weapon in a science fiction flick can shoot lasers, warp space, or spray hot grits, but no weapon, fictional or real, can have that many protrusions. You'd never get it into, our out of, a holster. Every branch and bush would tear it out of your hand. And a gun with a glass bulb as a functional unit? Are you kidding me? The reason the guns looked so awesome in Star Wars was because they were made from real guns. Many of them were made from, or based on, real, practical designs. The science fiction element was that they shot laser beams.
There's suspension of disbelief, then there's suspension of common sense. Not the same thing!
Rant over. Please return to your scheduled fawning. 8)
nor can you prove that the big bang was (not) a big explosion.
Don't confuse science with religion - the big bang theory is not taken on faith. It is a falsifiable scientific theory like any other. It most certainly can be proved wrong, if evidence is found that is not consistent with the model of the universe expanding from a denser state. Scientists spend a lot of time and money looking for such evidence, for example, the WMAP probe, which measures the properties of the cosmic microwave background to test the big bang theory. So far, the theory has checked out, but it would be dropped like a ton of bricks if we found, for example, stars much older than the expected age of the universe.
Science is about verifying and testing our theories, not about saying "why don't we just agree to disagree" so that we can all feel fuzzy and warm in our ignorance.
PS: While we're at it, I love the way creationists can't tell the difference between the big bang theory of the universe, and the evolution theory of biological life. Is schooling in America that bad that supposedly 'educated' people are having trouble differentiating between galaxies and animals?
I have a graduate degree in Physics, so I now occasionally browse wikipedia articles to read about topics that I learned about at University, such as the latest results in planet finding, quantum computing, and so forth. Many of these articles link to mathematics articles. Invariably, I find the maths articles to be beyond dense. They do not help me understand the physics article they were linked from better than I did before.
I think the problem with the maths articles can be summed up with your comment "cannot be [done] with the space it allows". You do realize that Wikipedia is not paper, right? There's no limit to the length of an article. It's like a programmer complaining that the reason their code is uncommented is because it makes the code too long!
Have you ever tried to understand a random piece of complex source code with single-letter variable names and no comments? It's hard, even for professional, full time programmers. The mathematics articles are like that -- hard to understand by anyone but the author.
Call me crazy, but when I read articles about geometry, I expect diagrams, not greek letters.
And before any mathematicians get defense about 'precision', or the required 'formalism', think of the introduction as commentary about the formulas, not a replacement of the formulas.
I think you'll find that it is you that doesn't understand what a snapshot could be. Take a look at ZFS, try it, and see if you think of snapshots the same way again. In ZFS, a snapshot can be promoted to a clone, which is a writeable copy of the original filesystem, sharing unmodified blocks using a copy-on-write algorithm.
This is increadibly powerful and useful. For example, a single master 'image' volume can have customizations added for specific purposes. This is useful in desktop deployment, iSCSI or NFS network boot, etc...
Would you expect a 'first class' writeable clone to have a name like 'dev/mapper/snapshotted-hda' or 'dev/hda.1'? Which one makes more sense? Why would the original have a special name, when the clone is identical?
It's this kind of narrow 'snapshots are throwaway' thinking that causes artifical limitations in APIs and operating system design that serve no real purpose.
As a Visual Studio user for many years, the most interesting comment I heard was from a fellow programmer who said that they prefer to turn off the tab-completion sometimes. The idea is that if you have no need to memorize the various functions in the standard library, you probably won't. He found that it increased his ability to plan ahead, reducing overall programming time. He still coded with tab completion on most of the time, but not always.
Another point of view is that Visual Studio automates a lot of things (such as GUI design), but this can limit creativity and flexibility, as coders will avoid doing things that the limited designers and wizards can't do. For example, the "Setup Project" wizard is handy, but it doesn't support optional sub-components that users can choose not to install, which means that coders who use it will most likely create "all or nothing" installers.
"Slow" in database terms is seconds. A table lookup on a sorted table will take at most 20-60 steps, and each is a single disk access. At a typical 5ms per access, that's no more than 100-300ms! A well designed index can reduce the number of steps required further. Only very naive database designs slow down significantly when the amount of data increases.
I have, and I agree. It's the one of the most depressing sights on this planet. Huge areas of dirty, gray, dead rock instead of brilliant colour and life.
- It takes as much as 10 seconds longer to open big docs sent in Office format (read: anything sent to you most people outside the company).
This is so trivial it's not even worth mentioning.
How is that trivial? I evaluated Open Office at my company that does software rollouts across a total of over 10,000 desktops, and among other things, one of the showstopper problems with OO was that it took up to 30 seconds or more to load a single page Word document on a 3.0GHz P4!
That's not a "slight performance issue"... that's abysmal. There are companies out there with 50MB+ documents. Can users be expected to wait.. what? Hours?
Slashdot Mirror
That's marketing bullshit.
I'm one of the people who runs Windows 2008 Server on a laptop, and I can tell you now, it's Vista with a higher retail price-tag.
Microsoft has been releasing server builds that are virtually identical to the desktop editions for years now. Windows XP 64-bit uses Windows 2003 service packs. Windows 2008 uses Vista drivers, the server editions have the DirectX gaming APIs, the workstation editions can serve file and web pages, etc...
The only difference in server is different initial default settings, some additional modules (such as Active Directory), and a half-dozen tuning parameters, most which are set through the registry anyway.
Even Active Directory will run on a desktop OS! It's called Active Directory Application Mode (ADAM).
This snipped caught my attention:
...construct a solution tailored to a specific environment...
The thing is, I'm a systems integrator, and as a part of my consulting job I have worked on small and large projects at something like 100 different companies and organisations of all sorts. Do you know what I discovered?
Companies all have the same needs when it comes to IT.
They all have the same need for home directories, application deployment, patching, security, backup, and the like. They all run the same kind of applications for the same reasons. It doesn't change as much as people think from company to company. So why should I have to reinvent the wheel every time? Why shouldn't I use a pre-packaged solution which does exactly what everyone needs? Isn't that a huge waste of time and money?
Notice that the Munich project ran over budget.
This is why Microsoft is so hugely successful. They realised that people don't want to "do it themselves" unless it's a hobby.
http://www.leekspin.com/
It makes my brain juices fizz.
I think the thing with teaching is to look at different approaches taken by other people, and use the method that actually produces results. I saw teachers try several different techniques at high school and university level, and they all failed miserably, except for one guy who was wildly successful. He was actually tasked with teaching us Haskell, which is an abstract pure functional language that most people "don't get". Despite that, he successfully taught it to the majority of students in his class, who then couldn't learn Basic or C from other teachers in the subsequent semesters.
His technique was simple. He set up a digital projector with his laptop so he could sit and face the class, and projected the development environment onto the screen. He then started coding really simple problems ("add all the numbers in a list") and running the example code, often through the debugger.
With almost every problem, he'd deliberately make a mistake (leave out a bracket, or not increment the counter in a loop, etc...) and demonstrate how the program failed, how to interpret the error messages or debug it to find the source of the error, and the result of fixing the error. He'd involve the class, asking the students to 'spot the error'.
I could already program when I started his class, so it was fascinating to watch him teach. What I found is that he was teaching exactly the things that I found difficult when I first started out:
- It's hard for beginners to go from a compiler error message to a fix. I still remember the Borland C++ error message "Cannot assign value to rvalue". It took me a day to figure out that I had an "=" instead of an "==" in an expression.
- Writing programs is relatively easy to learn, but learning how to "mentally execute" a program step-by-step to debug it, or use a debugger, is hard. Many beginner programmers just assume that it'll do what they meant, not what they actually wrote. Writing correct programs is much harder.
When I was studying computer science at university, I had read about all these fancy cryptographic techniques, and I imagined that banks were these encrypted, firewalled fortresses of IT security, monitored by the most competent, most vigilant administrators.
I was very wrong.
Let me tell you about my experience of IT security in banks.
A couple of years ago, I was sent to one of Australians largest banks. I was there for a 1 week engagement to install the latest virus scanner software on some servers. It sounded like a great opportunity to have a look at some high-end systems and see how they were managed. So I turn up in the morning, and start unpacking my laptop, when the project manager warns me:
"Don't plug your laptop into the network. We have to make sure you have the latest patches and AV first."
I fully understood his position, of course, they couldn't just let some random guy plug some a laptop into the network. It was a bank after all, security matters. I was Wrong. He corrected me:
"Oh no.. that's for your own protection! There's hundreds of viruses on this network, if you plug an unpatched machine into it, it'll be infected in seconds."
I was stunned. He wasn't even joking. I did plug my laptop in (which was well patched), and ran Ethereal for a few minutes, during which time I saw several viruses attempting to hack my machine. It was incredible. I've never seen that kind of attack rate anywhere, and I've been to large, unfirewalled university networks and school networks.
In fact, I didn't even really need to plug myself in. There was a WiFi connection available, with an easily recognisable SSID (the name of the bank). Of course, it was unencrypted, unsecured, and plugged directly into the desktop LAN.
Next, I got a tour of the data center, which was an eye-popping experience in itself. The bank had recently invested in fancy new retina-scanning door locks. It looked like it was straight out of a James Bond movie. However, it was taking too long to program in every person who needed access into the system, so they had simply propped the door open with a bucket. The inside of the room was just as scary. I walked past DOS machines, Windows 95 "servers", and I saw at least one NT 3.x machine. This was in 2005.
Eventually, I got around to planning the AV software upgrade. Except it wasn't. It was first-time-install, because the majority of their servers had no AV. The amount of work required to verify compatibility during for a the rollout was deemed too expensive, and I never did get to install the AV software. They did buy the licenses though, so it's entirely possible they installed it themselves. It's possible, but I wouldn't bet on it.
You might find this interesting (or might not).
As a preface, "How much physics do you know?": A lot.
Anyway, there was a paper in Astrophysical Journal a few years back, in either 2005 or 2006 from the University of Victoria. They got nice rotation curves for galaxies just from general relativity, without invoking dark matter. Kind of neat.
Mike.
I know a bit of physics too, having studied it at University, and Dark Matter always gave me the impression that when Physicists computed Galactic rotation, they did the typically lazy "good enough for Physics" mathematics and simplified -- they used Newtonian mechanics. I have not seen even one paper or article mentioning relativity. Not surprisingly, at the galactic scale, Newtonian mechanics doesn't quite work out. Soo.. wait for it.. what's the correct answer? There must be a mysterious, magical force out there! It can't possibly be that we used the equations that are generally known to be inaccurate. No sir. No way. This, err.. inaccuracy is caused by something I just discovered! Can I have my Nobel now?
The paper the parent article mentioned is:
General Relativity Resolves Galactic Rotation Without Exotic Dark Matter
http://xxx.lanl.gov/abs/astro-ph/0507619
30 TB per night sounds like a lot, but 1.5 TB drives are about AUD 350 each, retail. By 2016, I'd expect vendors to have released at least a 10 TB hard drive at that price point, and I wouldn't be surprised if we're using 30 to 50 TB drives.
So it all boils down to about $1000 per night of operation, or about $350K per year. Not exactly expensive for a science project. A single mars mission costs about $300M, but this telescope would generate more discoveries. That's not even considering that storage costs would continue to drop over the lifetime of the telescope, so the eventual total cost may be less than $100K per year. That's the salary of just one person!
Stop quoting nonsense you heard from your grandpa.
Film is a terrible archival medium, except for maybe silver based black and white film. It fades, the color changes, is easily damaged, and the original degrades when copied. George Lucas has spent $millions carefully restoring the archived Star Wars films, and they're a lot less than 50-60 years old. Film over 50 years old usually takes heavy processing to be even watchable.
On the other hand, digital archives are trivial to copy losslessly, so there's no need for any physical media to last for the length of the archival time.
Unfortunately the parent post is right - space colonization in the foreseeable future is unlikely for many reasons that somehow seem glossed over by the "true believers".
Lets face the cold hard reality of space - it is both cold and hard. There's nothing out there but rocks. Nobody wants to live on cold hard rocks. Some people might go there for science, or out of the curiosity of a tourist, but nobody will ever want to make a life there.
I'm sure of this because people already have the opportunity to go live on cold hard rocks that are far from civilization, right now, but don't. There are huge, unclaimed tracts of rocks right here on Earth. Most of Canada, Russia, and Antarctica is an uninhabited wasteland, despite massive overpopulation in other parts of the world. Heck, most of the continent of Australia is an uninhabited desert, so there's also a choice of warm hard rocks, if that's your preference.
Compared to any other place in space, these places aren't even that uninhabitable. There's usually some water, a breathable atmosphere, soil, gravity, real time telecommunications, and resources and equipment can be brought in a cost of mere thousands of dollars instead of billions. Yet despite these manifold advantages over space colonization, there's no popular demand for massive government spending to colonize these places. Why not?
Sure, it's not glamorous, but we can do it right now! We could, if we wanted to, colonize Antarctica. It wouldn't even be that hard, all the technology is available right now. We could move a billion people there if we had to. Does anybody want to though? Do you? Would you, right now, give up where ever you are, with your job, friends, family, and go live in a place like that? Or if you don't like the cold, you're welcome here in Australia! The desert has some really cheap land. You can buy a place the size of a small American state if you want to. Oddly enough, most of tens of thousands of immigrants that come to this country every year go to live in the larger cities. Not a huge demand for desert living for some strange reason.
Now let me put it this way - Mars is just like Antarctica, but much colder, much more remote, much harder to reach, much harder to come back from and there's no atmosphere. Not to mention that the return ticket for a family sightseeing tour these days is $1 Trillion*.
Living on Mars is totally irrational wishful thinking. Unless some miracle occurs like the sudden invention of cheap wormhole generators or anti-gravity or some similarly quick and easy way to get about becomes available, I just don't see it, and we can't make plans where step #1 depends on a miracle. Even given some cheap magic space transport device, every destination is still "Not Earth". Step #2 will be that anywhere we might want to go is instant-death-to-the-unprotected. Every little detail is lethal. Did you know that due to the lack of weathering, the dust on the Moon is microscopically jagged and razor sharp? It'll cut your lungs up if you breathe it in for any length of time. Just look up "silicosis" to find out all about the joys of that particular ailment. In comparison, the Australian Outback is so inviting that even the dust comes in a lung-friendly rounded format, but there's still places you can go where the next nearest person is 100km away. I don't see that changing any time soon, so why would anyone expect space to be colonized first?
* An estimate only, one that NASA keeps revising upwards. Before betting on cheap space travel, wait for the prices to actually start dropping first.
While it is true that Windows Server 2008 is almost exactly the same as Vista SP1, down to the hotfixes and drivers, the tangible difference is really a bunch of compiler macros and flags that Microsoft charges hundreds of dollars for.
I run Server 2008 with the "Desktop Experience" pack as a substitute for Vista on my work laptop because of a bad experience I once had while doing a demo for a customer on an XP laptop - I had developed a simple ASP.NET website and was making a demonstration when one of the users had managed to produce a "HTTP/500" error. It was incredibly embarrassing to have my supposedly "highly reliable" system lock up after just a few clicks. It took me days to figure out that the "crash" was caused by a completely artificial limitation introduced by Microsoft into XP to differentiate it from their Server line - one of the TCP/IP connection limits was the culprit. I had never noticed it while developing, because loopback connections are not affected.
So now I run an MSDN licensed Windows 2008 as a "workstation" OS so that I can avoid the Microsoft Marketing Department's deliberately introduced bugs, leaving only the plain old technical bugs, of which there are thankfully fewer than some previous MS operating system releases.
If Microsoft made cars, the replies to a question such as this would inevitably be "I want a car that goes 300 miles per hour, 5000 miles to the galon, has room for all the family, and.. and looks awesome! Yeah! With bells ..and whistles!".
Back to reality. What I would like from Microsoft is much simpler - I want them to do what they've delivered in the past but failed to do so recently. Microsoft's strength has always been consistency and compatibility. For example, virtually every platform called "Windows" has support for both Batch and VBScript automation. This is great, if as an admin, I want to write a script, I can. No problem, it will just work.
Recently they've introduced Powershell, which has made even Linux zealots a little jealous. Except it's not installed by default. That's right, in a 10GB base install of Windows Server 2008, they couldn't spare the 2MB scripting component. It couldn't make it in there. No room... or something. This means that unless a server requires Powershell (eg: because it runs Exchange 2007), it probably won't have it! This makes Powershell an order of magnitude less useful than it should be. Of course, you can install it with another script (a batch file, for example!), as long as you shove the DVD into the server.
So... you're a contractor, you're at a new site you haven't seen before but you know they have Windows 2008 on all their servers and you've been asked to bulk-reconfigure a bunch of servers on a deadline. You plan to use your Jedi PowerShell skills and your psexec bulk automation wizardry, don't you? Try again. It's back to Batch files for you, like everyone else.
This is why end-user ideas like "componentized operating systems" are stupid. Unless you're developing for the embedded market, shaving 1MB off a 10GB base install is just frustrating admins and software developers for no tangible gain. If a feature is not guaranteed to be installed, developers have to assume it isn't and hence can't depend on it or use it. Most users don't even know where their install DVD is, if they even have one, so they can't fix any missing components required by a new piece of software.
And Microsoft really has to drop the ridiculous "security through signature" crap. You have to sign Powershell scripts now. Think for a minute how ridiculous that is. You can run unsigned executables, binary blobs with any possible content at all, but not a script? How does that makes sense? The first thing every admin does is turn that feature off, resulting in a hassle and no additional security at all. Scripts, like everything else, are already restricted to whatever the user can do, they need no additional security. Not to mention that no-one would be stupid enough to run an script from an untrusted third-party without glancing through it first!
I just tried this with Mozilla 3.0 beta (Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.9b4pre) Gecko/2008021607 Minefield/3.0b4pre) and the score was 60/100. Not bad.
However, it doesn't look like the reference rendering. There's no colors, there are red blocks in a few places, and there's a purple "X" as well.
That used to be true, at least for hardware-level corruption.
See: http://en.wikipedia.org/wiki/ZFS
Also, I'm betting they were using consumer-grade hardware, which is notorious for skipping the 5$ extra chip needed for ECC.
I hate this 'rule of thumb' that people keep trotting out that we only use x% of software, for some low value of 'x'. That's simply not true, so stop bringing it up like it's a valid argument for anything.
I keep hearing this bullshit if from Unix zealots, from people flogging 'thin' or 'web 2.0' products, and from Luddites that are 'perfectly happy' running WordPerfect 5.1 on their OS/2 machine.
Lets think of a simple scenario. Imagine a fictional company MiniSoft Software that makes a word processor. They advertise that their program has 100 features! Of course, you know that most users will only use about 10% of that most of the time, and maybe an occasional 1% rarely. So why have the other 89 features in there? Most users won't be using it!
What this kind of oversimplified 'analysis' misses is that that '1%' extra is different for every user. Glenda in marketing might use the 'mail merge' feature once a month. The payroll officer might have to use the database integration feature. The warehouse manager might be using the barcode printing. The international sales office might use the Unicode multi-lingual features.
Once you add up all of those '1%' pieces, all too often, you end up with... 100 features or so. This is why MiniSoft Office is so 'bloated'. Because somewhere, out there, there's someone who uses the macro functionality, or the right-to-left text input, or the dynamic forms, or... something. It's not bloat... it's what users expect from their software -- that the same consistent product be useful for all of the staff in an entire business.
So to reiterate, just because YOU only personally use the "bold" and "italic" buttons on the toolbar doesn't mean that someone else can get by with only those two buttons.
Get used to it, because software is only going to get bigger and more 'bloated', not less.
There is no user intervention required: They could be opened by a cord attached to the exit, just like the airforce uses for large paratrooper jumps.
And are dead passengers much better for liability? This is exactly like the car manufacturers dragging their feet on installing seatbelts. There's no excuse. If the parachutes could save lives in some circumstances, they should be installed. Life jackets on the other hand are useless. Planes hitting the water 'at speed' invariably disintegrate and kill all aboard. However, many 100% fatal accidents have been over land, with enough available time to get the passangers out. Parachutes could have saved hundreds of lives.
Think of it this way: In world war two, tens of thousands of allied soldiers were thrown out of planes, in the dark, while getting shot at. Most of them survived. Some people broke bones, some didn't have their chute open, but by and large, most people survived.
The hardest part to having people jump from a jumbo is getting 300+ people to put on their parachutes in a reasonably short period of time. That's the only thing I can think of that would be a stumbling block. Everything else can be solved.
I wish more artists would embrace realistic fantasy.
It sounds like an oxymoron, but it's the difference between a movie only kids could enjoy, and something adults would want to go see too.
The first thing that struck me about those pictures is that nobody would ever, ever, ever use one of those contrived contraptions in a battle. A weapon in a science fiction flick can shoot lasers, warp space, or spray hot grits, but no weapon, fictional or real, can have that many protrusions. You'd never get it into, our out of, a holster. Every branch and bush would tear it out of your hand. And a gun with a glass bulb as a functional unit? Are you kidding me? The reason the guns looked so awesome in Star Wars was because they were made from real guns. Many of them were made from, or based on, real, practical designs. The science fiction element was that they shot laser beams.
There's suspension of disbelief, then there's suspension of common sense. Not the same thing!
Rant over. Please return to your scheduled fawning. 8)
Don't confuse science with religion - the big bang theory is not taken on faith. It is a falsifiable scientific theory like any other. It most certainly can be proved wrong, if evidence is found that is not consistent with the model of the universe expanding from a denser state. Scientists spend a lot of time and money looking for such evidence, for example, the WMAP probe, which measures the properties of the cosmic microwave background to test the big bang theory. So far, the theory has checked out, but it would be dropped like a ton of bricks if we found, for example, stars much older than the expected age of the universe.
Science is about verifying and testing our theories, not about saying "why don't we just agree to disagree" so that we can all feel fuzzy and warm in our ignorance.
PS: While we're at it, I love the way creationists can't tell the difference between the big bang theory of the universe, and the evolution theory of biological life. Is schooling in America that bad that supposedly 'educated' people are having trouble differentiating between galaxies and animals?
I have a graduate degree in Physics, so I now occasionally browse wikipedia articles to read about topics that I learned about at University, such as the latest results in planet finding, quantum computing, and so forth. Many of these articles link to mathematics articles. Invariably, I find the maths articles to be beyond dense. They do not help me understand the physics article they were linked from better than I did before.
I think the problem with the maths articles can be summed up with your comment "cannot be [done] with the space it allows". You do realize that Wikipedia is not paper, right? There's no limit to the length of an article. It's like a programmer complaining that the reason their code is uncommented is because it makes the code too long!
Have you ever tried to understand a random piece of complex source code with single-letter variable names and no comments? It's hard, even for professional, full time programmers. The mathematics articles are like that -- hard to understand by anyone but the author.
Call me crazy, but when I read articles about geometry, I expect diagrams, not greek letters.
And before any mathematicians get defense about 'precision', or the required 'formalism', think of the introduction as commentary about the formulas, not a replacement of the formulas.
I think you'll find that it is you that doesn't understand what a snapshot could be. Take a look at ZFS, try it, and see if you think of snapshots the same way again. In ZFS, a snapshot can be promoted to a clone, which is a writeable copy of the original filesystem, sharing unmodified blocks using a copy-on-write algorithm.
This is increadibly powerful and useful. For example, a single master 'image' volume can have customizations added for specific purposes. This is useful in desktop deployment, iSCSI or NFS network boot, etc...
Would you expect a 'first class' writeable clone to have a name like 'dev/mapper/snapshotted-hda' or 'dev/hda.1'? Which one makes more sense? Why would the original have a special name, when the clone is identical?
It's this kind of narrow 'snapshots are throwaway' thinking that causes artifical limitations in APIs and operating system design that serve no real purpose.
Finally!
I've been waiting for someone to make this point on Slashdot eloquently and be modded up.
As a Visual Studio user for many years, the most interesting comment I heard was from a fellow programmer who said that they prefer to turn off the tab-completion sometimes. The idea is that if you have no need to memorize the various functions in the standard library, you probably won't. He found that it increased his ability to plan ahead, reducing overall programming time. He still coded with tab completion on most of the time, but not always.
Another point of view is that Visual Studio automates a lot of things (such as GUI design), but this can limit creativity and flexibility, as coders will avoid doing things that the limited designers and wizards can't do. For example, the "Setup Project" wizard is handy, but it doesn't support optional sub-components that users can choose not to install, which means that coders who use it will most likely create "all or nothing" installers.
"Slow" in database terms is seconds. A table lookup on a sorted table will take at most 20-60 steps, and each is a single disk access. At a typical 5ms per access, that's no more than 100-300ms! A well designed index can reduce the number of steps required further. Only very naive database designs slow down significantly when the amount of data increases.
I have, and I agree. It's the one of the most depressing sights on this planet. Huge areas of dirty, gray, dead rock instead of brilliant colour and life.
- It takes as much as 10 seconds longer to open big docs sent in Office format (read: anything sent to you most people outside the company).
This is so trivial it's not even worth mentioning.
How is that trivial? I evaluated Open Office at my company that does software rollouts across a total of over 10,000 desktops, and among other things, one of the showstopper problems with OO was that it took up to 30 seconds or more to load a single page Word document on a 3.0GHz P4!
That's not a "slight performance issue"... that's abysmal. There are companies out there with 50MB+ documents. Can users be expected to wait.. what? Hours?
No?
... then users will keep using MS Office!
True, it installs a virtual network card driver, but this isn't at all unusual, as VMWare and Virtual PC do this too.
He's just trying to plug a Virtual Machine solution that isn't anything at all like CoLinux.