Now, complain about their products, complain about their service, complain about the prices, and they'll listen. Quit buying their products, and they'll listen. Hurt them in their pocket book, and they'l listen.
You're right about the second two, but wrong about the first. Plenty of people have complained about their products, price, and server, and Microsoft does nothing because people CONTINUE TO PAY. "Software Assurance" didn't get rolled back just because people were complaining about it, because they all still paid for it, regardless of their complaints. The only way you get Microsoft to listen is to tell them you are not going to buy their products. That gets them to sit up and listen, and possibly reduce the price just to keep you.
Bitching and moaning about Microsoft without action acomplishes nothing. If you don't like Microsoft's products, prices, or business practices, the most effective thing to do is take your business elsewhere, even if it hurts you in the short term.
I've seen a large number of hard drives with a large number of defects. Controller failures are one of the most annoying. The ones where the drive just suddenly dies are tolerable. The ones that are dangerous are the ones that corrupt data, but otherwise work fine (e.g. the onboard RAM is defective). It's the kind of error that destroys data without you ever really knowing it, especially if the drive can write fine, but data stored within it's cache is what's corrupted. It can destroy a RAID set up before you even know what's happened.
I believe the primary reason for controller board failure are simple defective parts that weren't defective enough to fail testing. Eventually the part will break, and the best power and air conditioning will only delay the inevitable. The other controller failure type is a firmware flaw. Ask the people who bought IBM's 75GXP and 60GXP drives about this. It is believed the drive's firmware could not handle thermal recalibration properly, and eventually failed because of this.
SMART is great for some failures... Particularily the ones that are caused by a slow degredation of the media or drive mechanics. Unfortunately, there's plenty of failure types that can't be predicted by SMART error counters. Such as drive controller failure, head crashes, or power spikes. When SMART works, it's great. A chance to salvage your data, possibly without any data loss. When it doesn't work...
Excuse me? Reliability of backup to disk is... higher? You may have had some bad experiences with tape drives (and there are some very bad tape drive technologies out there), but tape drives are very reliable beasts. 4mm DAT drives, DLT, and LTO drives all have a reputation for being very reliable. Add to this the fact that tapes are designed to be removed, and are durable enough to handle minor falls. Most hotswap removable hard drive bays are not designed to be swapped out every day, 261 days a year or more. The LTO drive I use has a rated archival life of 30 years, although if I backed up to one particular tape every day for a year, I'd have to replace it.
The other part of this is that when tapes are going bad, you have some advance warning. You start seeing hard errors while writing to the media, and will have a chance to order new media before the tape is completely bad. With a hard drive, you don't know it's bad until it's too late, and is developing bad sectors.
Just because you had a QIC, Travan or iomega Ditto drive, and it was junk, doesn't mean all tape technology is unreliable.
Oh, please. A managed runtime is not a magical security bullet. In the case of Internet Explorer here, these are not the buffer overflows, off-by-one or signedness errors that a managed runtime could ever protect against. These are simple security design errors. Microsoft wanted to show how great their IE engine was and implemented security zones so that local HTML-only applications could exist using the engine. They are being burned by this, now, as people find new ways to turn the higher security 'Internet Zone' into the lower security 'My Computer' or 'Trusted Sites' zones.
Of all programming errors, buffer overflows, off-by-one, and signed mistakes are some of the easiest spot and to fix. Other errors, like SQL injection, privledge separation, races and the dozens of other errors that can cause crashes, security vulnerabilities, or denial of service attacks, can not be protected against by a managed language because they're outside the scope of the language itself.
Okay, maybe I'm just foolish, but how does one go about 'measuring' political bias? Is there a geiger counter that tells you how leftist someone (or a story) is? Is there a form of litmus paper that turns blue when applied to conservative stories, but red when applied to liberal biased ones?
I skimmed through that study, and I think the methods are a little loony. Determining how biased a news source is based on which political 'think tanks' they quote most often? I have a pretty low opinion of these 'think tanks' to begin with (they seem like nothing more than hired mercenaries for their corporate sponsors 9 times out of 10) that I think using them as any kind of meaningful metric is wrongheaded. You might as well try to measure liberal or conservative bias by measuring the degrees of separation from Ann Coulter, for crying out loud.
Don't worry. I'm sure Vivendi's management can kill off Blizard before too long, just like EA has managed to kill off all their decent developers so they can put out Madden 2009, now with freshly cut grass!!!
Have games really become so complex that independent dev houses no longer work, or is it that publishers have this idea in their head that they'll make more profits by gobbling up independant game houses? This consolidating of the game biz is destroying it, IMO.
More secure? For certain definitions of secure, perhaps. Windows TS allows you to use NT security and permissions to lock down a user from changing what he/she shouldn't. RDP5 also includes encryption, so everything you send isn't clear text over the network. For an equal number of users, Windows terminal server is likely to be more space, memory and CPU efficient. Machine stability should also be better on the Windows Terminal Server because of the NT kernel (be honest, Win9x crashes... a lot). The only benefit I can really see to Win4Lin TS would be compatibility for apps that do not run on NT kernels, licensing (less CAL nonsense), and the fact that Win4Lin sessions are probably more isolated from each other than TS sessions are.
It is rather funny the accusation works both ways, isn't it? Most people who make such accusations for either side have usually have either only seen the other method, or sat down with it for approximately five minutes, decided they didn't like it, and became the all-knowing expert on all things development related.
Personally, I've seen benefit in both paths. vi/make/etc are wonderfully lightweight tools, and in the time it takes to load VS.NET, you can sometimes have the changes made that you needed to make, have it recompiled, and be testing it. On the other hand, developing GUIs in code alone really sucks. It's a lot of trial and error just to get the spacing right. And a debugger with a GUI is a heck of a lot easier to work with than gdb (especially considering the only command I ever remember is 'bt'). When debugging multithreaded applications, the utility of an IDE-integrated debugger just goes up. In terms of actually writing code, I find it's just a completely different process. In an IDE, you'll just be using that one program the entire time. With vi/emacs/make, you'll end up switching between number of different VTs.
To make matters worse, on Thunderbird, it says "Select All Ctrl+A", but all ctrl-a does is bring you back to the start of the line. Alt-A is the select all command.
One has to wonder about how much longer Borland can hold out. Their development tools have always been top notch, but they've always lost out to Microsoft because whenever Microsoft comes out with new APIs, they're always available first for Visual Studio. Borland users either have to import them themselves, find someone else who's done it for them, or wait for the next version.
I'm really not sure how they're still around today, but I'm grateful that they are. They're still building interesting technology, today.
Re:Take a look at the accessories at the bottom. .
on
3D Linux Laptop Available
·
· Score: 3, Informative
And $50 Logitech mouse, that you can buy direct from Logitech for half that, and probably even less at any retail store without even needing a sale. Or the $150 3COM 802.11b/g PC card that you can pick up for 60% less anywhere else. Or even their 60GB hard drive upgrade for $500, that I can get for less than half price in Canadian funds, ignoring the exchange rate, here and it's a faster drive, too!
For anyone who's buying from them, stay away from the accessories. Mice are universal, PS/2 or USB. Laptop hard drives are universal, and the only thing you might need to worry about is height (9.5mm or smaller?). Laptop memory may not be completely universal, but it's pretty easy to findcompatiblestuff. All or virtually all external USB storage devices are compatible with Linux.
60 watts for your laptop? I think you're looking at the wrong figure. That's probably the power output to your laptop, not the power input. On my laptop, the power transformer does 75W (15V @ 5A) out, but 120-144W (1.2A @ 100V or 0.6 @ 240V) in. It's between 63% and 52% efficient by those numbers. My laptop may consume less than half the power my desktop does, but it comes at a price. A poor keyboard, poor pointing device, slow hard drive, mediocre video chip and a slow response rate on the high resolution LCD.
It was one of the first games that used positional sounds as an integral part of the game, [...] that spawned the stealth genre which now also includes games like Metal Gear Solid and Splinter Cell. [...] This new genre is very successful at the moment, with sequels for all three games mentioned coming out this year - Thief 3: Deadly Shadows, Metal Gear Solid 3 and Splinter Cell: Pandora Tomorrow, all of which have received good to excellent reviews.
There's also the Rainbow Six series, the latest of which is Raven Shield. A great game as long as you don't mind being very, very careful. One or two shots will probably kill you, and spraying bullets at an enemy is foolish. It's first person, unlike the rest of the games you posted. Hitman 2 was another game that makes you use stealth. The game XIII put an interesting spin on it, too, as you could not only hear your enemy, but also "see" their movement on screen.
And what do you see people talking about in this story.... A 6 year old game with stinky sound to begin with.... CounterStrike. I wish some of these counterstrike people would get off their asses and see there's an entire world of PC games just waiting to be explored. After going to a large LAN party event with over 800 people, and only being able to find ~5 people who were even a slight bit interested in playing Raven Shield, and maybe a 15-20 people who were willing to play BF1942, I have to say I'm a bit bitter as CS.
One? linuxthreads is a subset of POSIX threads, and is part of glibc now. There's three different implementations of threads on Linux, the original linuxthreads(1:1), ngpt(M:N), and nptl(1:1). They might be implemented in different libraries, but they're just one API -- POSIX. Of course, maybe you mean Apache's APR, wxWidget, and the other platform abstraction libraries?
I would like to thank you for putting up that website. It's enlightening if somewhat depressing. If the majority of respondants had even replied in the way Mike Redmond had, saying "I don't know", I would've had more respect for most of them. I'm a bit surprised with the Conservative party's responses, as I would've thought they'd be doing centralized control, too.
A bad situation is brewing and most people don't even realize it yet. As Canadians, we pay a levy on blank media for the privledge of private copying, yet, the recording industry is starting to put copy protection on their CDs, and is lobbying to get that legal protection established for their copy prevention technology. So, while we'll legally be able to make a copy, we'll be technologically prevented from doing so, and legally prohibited from beaking the technological protection. The recording industry gets to have their cake and eat it, too. Whenever I need an example of why legal protection for TPMs is a bad idea, this is it. The CD-R levy has a lot of things going against it, but it's probably a fair compromise, but if it gets backed up by TPMs, then its fairness goes down the drain.
Another point worth to be noted is that, under Un*x, the DLL Hell is a non-issue, as we've had libraries versioning since day 1. So, I might as well install multiple versions of a library, and yet do not have the need to recompile an application.
I wish that were true. I really, really wish that were true. Too many people do not understand the correct way to determine the version numbers on their libraries, and instead fix them to whatever release of the package is. Ask the people who packaged KDE for Debian why this is a good idea, and I'll bet you'll get an earfull.
I suppose it's just Murphy's law. There's two ways for programmers to interpret the version number on.so objects, so naturally they choose the disasterous way.
My point was merely that in any Real program, it's very likely the garbage collector will be called. In these benchmarks, it's unlikely it's ever called. A performance-wise analogy to C++ would be to create the objects and never free them. It's bad programming practice, but it would make the C++ and Java versions just as dishonest about their speed. Although, it would be a valid benchmark for one part of the process -- object creation.
I think GC is a great idea that a good implementation of which could lead to increased performance in a lot of situations.
I'd agree, except for one thing. Most garbage collected languages seem to assume that one-size-fits-all. When I was writing some C# code, I found myself wanting a delete operator, not because I wanted to clean up memory, but because I wanted to expliticly say that "I am done with this object, please do not allow me to use it further" as a guard against some programming mistakes I might make. In a C++ program, I'd get an access violation as I tried to access already freed memory, but there was no way of getting this kind of runtime error in C#.
Unfortunately, after reading stuff from the designers of the language (mainly telling users, 'you don't need that'), I was disappointed with their attitude of academic snobbery. They had the same attitude when it came to macro expansion, or anything else that language designers fault for the collapse of civilisation.
Because it was for a really old version of excel (for Windows 3.x). GetSystemTimer was a win16 API call, along with SetSystemTimer, KillSystemTimer and a handful of other APIs. This was covered in the book, "Undocumented Windows", and is where the 'common knowledge' about Microsoft using undocumented APIs in Office came from. I won't comment on whether Microsoft has become better at it or not, but I will say that Microsoft Office is often the bringer of new APIs to Windows even today.
Okay, after much trial and error, I've concluded that you must be posting in "Plain Old Text" mode. The "HTML Formatted" mode I normally use changes multiple spaces to a single space, and filters out , even in tags. The end result is that is just plain broken in HTML mode, so it's a bug in slashdot.
Slashdot Mirror
Bitching and moaning about Microsoft without action acomplishes nothing. If you don't like Microsoft's products, prices, or business practices, the most effective thing to do is take your business elsewhere, even if it hurts you in the short term.
I believe the primary reason for controller board failure are simple defective parts that weren't defective enough to fail testing. Eventually the part will break, and the best power and air conditioning will only delay the inevitable. The other controller failure type is a firmware flaw. Ask the people who bought IBM's 75GXP and 60GXP drives about this. It is believed the drive's firmware could not handle thermal recalibration properly, and eventually failed because of this.
SMART is great for some failures... Particularily the ones that are caused by a slow degredation of the media or drive mechanics. Unfortunately, there's plenty of failure types that can't be predicted by SMART error counters. Such as drive controller failure, head crashes, or power spikes. When SMART works, it's great. A chance to salvage your data, possibly without any data loss. When it doesn't work...
The other part of this is that when tapes are going bad, you have some advance warning. You start seeing hard errors while writing to the media, and will have a chance to order new media before the tape is completely bad. With a hard drive, you don't know it's bad until it's too late, and is developing bad sectors.
Just because you had a QIC, Travan or iomega Ditto drive, and it was junk, doesn't mean all tape technology is unreliable.
Of all programming errors, buffer overflows, off-by-one, and signed mistakes are some of the easiest spot and to fix. Other errors, like SQL injection, privledge separation, races and the dozens of other errors that can cause crashes, security vulnerabilities, or denial of service attacks, can not be protected against by a managed language because they're outside the scope of the language itself.
I skimmed through that study, and I think the methods are a little loony. Determining how biased a news source is based on which political 'think tanks' they quote most often? I have a pretty low opinion of these 'think tanks' to begin with (they seem like nothing more than hired mercenaries for their corporate sponsors 9 times out of 10) that I think using them as any kind of meaningful metric is wrongheaded. You might as well try to measure liberal or conservative bias by measuring the degrees of separation from Ann Coulter, for crying out loud.
Or punchcards, with an automatic random shuffler.
Who needs the cast from Star Trek (TOS) to act stupidly illogical when you've got slashdot?
Have games really become so complex that independent dev houses no longer work, or is it that publishers have this idea in their head that they'll make more profits by gobbling up independant game houses? This consolidating of the game biz is destroying it, IMO.
More secure? For certain definitions of secure, perhaps. Windows TS allows you to use NT security and permissions to lock down a user from changing what he/she shouldn't. RDP5 also includes encryption, so everything you send isn't clear text over the network. For an equal number of users, Windows terminal server is likely to be more space, memory and CPU efficient. Machine stability should also be better on the Windows Terminal Server because of the NT kernel (be honest, Win9x crashes... a lot). The only benefit I can really see to Win4Lin TS would be compatibility for apps that do not run on NT kernels, licensing (less CAL nonsense), and the fact that Win4Lin sessions are probably more isolated from each other than TS sessions are.
Personally, I've seen benefit in both paths. vi/make/etc are wonderfully lightweight tools, and in the time it takes to load VS.NET, you can sometimes have the changes made that you needed to make, have it recompiled, and be testing it. On the other hand, developing GUIs in code alone really sucks. It's a lot of trial and error just to get the spacing right. And a debugger with a GUI is a heck of a lot easier to work with than gdb (especially considering the only command I ever remember is 'bt'). When debugging multithreaded applications, the utility of an IDE-integrated debugger just goes up. In terms of actually writing code, I find it's just a completely different process. In an IDE, you'll just be using that one program the entire time. With vi/emacs/make, you'll end up switching between number of different VTs.
To make matters worse, on Thunderbird, it says "Select All Ctrl+A", but all ctrl-a does is bring you back to the start of the line. Alt-A is the select all command.
I'm really not sure how they're still around today, but I'm grateful that they are. They're still building interesting technology, today.
For anyone who's buying from them, stay away from the accessories. Mice are universal, PS/2 or USB. Laptop hard drives are universal, and the only thing you might need to worry about is height (9.5mm or smaller?). Laptop memory may not be completely universal, but it's pretty easy to find compatible stuff. All or virtually all external USB storage devices are compatible with Linux.
60 watts for your laptop? I think you're looking at the wrong figure. That's probably the power output to your laptop, not the power input. On my laptop, the power transformer does 75W (15V @ 5A) out, but 120-144W (1.2A @ 100V or 0.6 @ 240V) in. It's between 63% and 52% efficient by those numbers. My laptop may consume less than half the power my desktop does, but it comes at a price. A poor keyboard, poor pointing device, slow hard drive, mediocre video chip and a slow response rate on the high resolution LCD.
And what do you see people talking about in this story.... A 6 year old game with stinky sound to begin with.... CounterStrike. I wish some of these counterstrike people would get off their asses and see there's an entire world of PC games just waiting to be explored. After going to a large LAN party event with over 800 people, and only being able to find ~5 people who were even a slight bit interested in playing Raven Shield, and maybe a 15-20 people who were willing to play BF1942, I have to say I'm a bit bitter as CS.
One? linuxthreads is a subset of POSIX threads, and is part of glibc now. There's three different implementations of threads on Linux, the original linuxthreads(1:1), ngpt(M:N), and nptl(1:1). They might be implemented in different libraries, but they're just one API -- POSIX. Of course, maybe you mean Apache's APR, wxWidget, and the other platform abstraction libraries?
A bad situation is brewing and most people don't even realize it yet. As Canadians, we pay a levy on blank media for the privledge of private copying, yet, the recording industry is starting to put copy protection on their CDs, and is lobbying to get that legal protection established for their copy prevention technology. So, while we'll legally be able to make a copy, we'll be technologically prevented from doing so, and legally prohibited from beaking the technological protection. The recording industry gets to have their cake and eat it, too. Whenever I need an example of why legal protection for TPMs is a bad idea, this is it. The CD-R levy has a lot of things going against it, but it's probably a fair compromise, but if it gets backed up by TPMs, then its fairness goes down the drain.
I suppose it's just Murphy's law. There's two ways for programmers to interpret the version number on .so objects, so naturally they choose the disasterous way.
Unfortunately, after reading stuff from the designers of the language (mainly telling users, 'you don't need that'), I was disappointed with their attitude of academic snobbery. They had the same attitude when it came to macro expansion, or anything else that language designers fault for the collapse of civilisation.
Because it was for a really old version of excel (for Windows 3.x). GetSystemTimer was a win16 API call, along with SetSystemTimer, KillSystemTimer and a handful of other APIs. This was covered in the book, "Undocumented Windows", and is where the 'common knowledge' about Microsoft using undocumented APIs in Office came from. I won't comment on whether Microsoft has become better at it or not, but I will say that Microsoft Office is often the bringer of new APIs to Windows even today.
Okay, after much trial and error, I've concluded that you must be posting in "Plain Old Text" mode. The "HTML Formatted" mode I normally use changes multiple spaces to a single space, and filters out , even in tags. The end result is that is just plain broken in HTML mode, so it's a bug in slashdot.