Slashdot Mirror
AV-Test Deems Windows Security Essentials "Very Good"
CWmike writes "Microsoft's new free security software, Windows Security Essentials, passed a preliminary antivirus exam with flying colors, said independent and trusted firm AV-Test, which tested Essentials, launched yesterday in beta, on Windows XP, Vista and Windows 7. It put it up against nearly 3,200 common viruses, bot Trojans and worms, said Andreas Marx, one of the firm's managers. The malware was culled from the most recent WildList, a list of threats actually actively attacking computers. 'All files were properly detected and treated by the product,' Marx said in an e-mail. 'That's good, as several other [antivirus] scanners are still not able to detect and kill all of these critters yet.' It also tested well on false positives."
viruses, trojans, and worms, are all nice and dandy but what about malware? That is what most people have to worry about these days.
MS is lifting their game.exe
sudo mount --milk --sugar
Norton, Symantec and others have created an entire multi-billion dollar subscription based industry around virus protection for Windows. I wonder how they are going to react to this potential bomb for there business model?
It's interesting, but at this point can Microsoft really convince anyone that they are serious about putting out a quality product? I think that is there biggest problem here... PR.j I will admit I laughed when I saw the article, and it is Microsoft's reputation that made me laugh. Maybe it is good, but I am I really willing to give them the chance with something that important?
I can remember articles talking about Windows Firewall in the past as being pretty darn good too, yet it seems the first thing a tech person does is to deactivate these days.
Let's face it. If Microsoft was seriously competent about doing these "core" activities, would the 3rd party market be as big as it is?
In any case it will be interesting if they start shipping Windows with this pre-installed. Then maybe the manufacturers won't be so quick to bundle Norton/McAffee with their products, and THAT will be fun to watch.
With all that talent, resources, and internal knowledge they should have a slam dunk. Unfortunately I have a lot of distrust built up from over the years about what MS sticks under the hood. It will take many years of good reviews and endorsements before I feed comfortable that the MS AV does not give any special passes to iffy software from a MS partner, or that the MS firewall will correctly block things from going out when configured to if the originator is an MS component.
The Microsoft style is to solve problems by throwing a lot of people at it, and they use that strategy fairly well. Instead of simplifying the structure to where it can be reasonably dealt with by a small group of people, they are happy to make it big. For example, compare the number of system calls in the windows kernel with the number in the Linux kernel. Having so many more system calls means each internal refactor will have to take more into consideration, as well as requiring more testing, but it's ok, Microsoft is happy to throw lots of testers at it. The ASP.net model, which basically wraps a whole system around html/javascript to encapsulate it and make it easier for the average programmer was an amazingly man-hour intensive job, once again requiring lots of testing and many special cases, and yet Microsoft did it.
That operating style is especially well suited to AV software, because it is a job that can be easily broken up and handed out to different programmers, and catching all the viruses is a job that can be easily helped if you have a lot of programmers and testers. It makes sense that Microsoft would write good AV software.
Qxe4
There was talk about antitrust suits when Microsoft first included the TCP/IP stack in windows. Before that you had to go to another vendor.
It made life a LOT more easy once it was built into the OS.
I'm pretty sure the same thing will be true of AV software.
It's always been this way. Microsoft rests on their laurels until an upstart company starts making money at their expense. Between Mac, Linux, and the insane proliferation of general crapware, MS has a real image problem on their hands. Luckily for Microsoft, the best and the brightest can be wooed by the kind of money they're able to throw around. When they throw their top programmers at a job, the results are stunning, just witness the turnaround from early Vista to the current beta of Windows 7.
Sadly, the end result will be bad for consumers. Other security companies will be badly hurt by the release of this freebie, and MS will go back to sleep, leaving the security marketplace to stagnate like the pre-Firefox browser market stagnated.
Just another "DOJ fascist authoritarian totalitarian bootlicker" -- Zeio
TCP/IP should be in the OS - it is a resource management issue and is a hardware issue. If only a handful of apps used it, maybe third-party would be acceptable. But you don't want a third-party stack crashing the OS, so write it yourself and include it.
On the other hand, Anti-virus products shouldn't even be needed. MS should be able to write software with fewer holes in it. They have piles of static analysis tools, piles of research, and piles of other stuff. They just don't want to take the time and fix things (including testing), so they put wrappers like UAC around things instead of fixing it.
I've seen lots of bug reports ignored by MS just because it doesn't look like it's exploitable, only to have some crafty fool figure out how to exploit it. I can cause a stack overflow in Oracle 10.x drivers by sending a VALID openquery through a linked server. Runtime catches the error, but then it causes a crash in the error reporting because the stack is trashed. Currently it's a null refrence, but how hard would it be to turn a stack overflow into a server root hack? Not all that hard. But they won't fix it because the problem is in a third party module, and if that one is fixed the MS error disappears. I'm just saying these vulnerabilities are all over the place, especially since they have so much third-party code.
One person or company making the problem, and the solution to the problem, does not look good. Especially since MS only publically fixes holes they publically admit to. There might be piles of security problems no one else knows about, but MS AV might know to watch for suspicious behaviour that only MS knows about.
Giant virus outbreak because it's too expensive to patch a particular problem, or can't get it out fast enough, and only Windows SE customers are protected so everyone ditches Symantec and other AV and goes to Windows. It's not that far-fetched, and they might even do it that way by accident. When it's possible to have that kind of advantage and wipe out your competition in a single event like that, especially if it's unintentional, that's a problem.
Just saying, the mafia used to take protection money, but you were being protected from the mafia. Problem and solution should be from different sources. Therefore your analogy is invalid, same as if my hair were a bird.
I imagine it will be a vary hard case to make since really all they are trying to do is fix their broken OS.
How is releasing anti-virus software fixing their "broken OS?" Are you implying that a non-broken OS is completely immune to viruses and malware or are you just spewing typical anti-Microsoft vitriol?
Oh ok... Thought so.
Similes are like metaphors
Microsoft should not be making antivirus software. It should be fixing its vulnerabilities.
The OS has many fundamental problems, some of which cannot be resolved without redesigning the core internals which would render all older software incompatible any newer version of the OS. This sort of problem was identified long ago, but it was decided that the cost of change would be too great, the burden on third party software vendors too heavy and ultimately, it would be too slow to adopt and migrate for all users. And the longer they wait for this eventuality, the more expensive and prohibitive it becomes to make such important changes.
If this sounds like the U.S. moving from Imperial measurements to the globally accepted Metric system, you wouldn't be alone in this observation.
Microsoft still cannot fix the "stupid user" problem but there are many things they could fix if they had the balls to do it. And they could take a page out of Deep Freeze's playbook and create a system where the user must first unlock the system before they can install anything. But perhaps the similarity to the adoption of the metric system doesn't stop here. Perhaps there will come a point at which everyone will move on to another system leaving the "imperial" one behind... well I can dream can't I?
Too bad apache proves you wrong.
This is an old, tired and false argument. If any OS had the holes windows does it would get exploited like mad.
Try using it.
1.
A shell that uses objects is asinine.
It looks like a bunch of java idiots tried to make a shell. If I wanted objects I would use a programming language, this is supposed to be scripting.
2. No ssh, lame.
3. does not support anything like authorized_keys.
Any computer consultant worth his salt won't get drawn into silly squabbles over OS/platform/software/language/etc., and will recommend the *best* solution for the client. Don't ever let bigotry blind you...
I describe myself as a dyed-in-the-wool Unix proponent (24 years now), but I run Windows on my desktop machines, and have recommended Windows on many occasions, including some large-scale Fortune 20 deployments, where it made more sense. (For servers, I avoid Windows unless the app environment really needs it or runs markedly better there, but there are still a good number of those situations. Given my druthers, I design new systems around open source technologies, mostly because of the lifecycle cost savings. Auditing all those licenses is a non-trivial cost and PITA, not to mention acquiring them in the first place - and avoiding licensed software makes leveraging cloud computing *much* easier...)
Windows certainly has its faults, and I'm a big critic, but it also has its place, and for a good number of things (even some server-based things), Windows is the best choice - sometimes by a good margin.
"The future's good and the present is nothing to sneeze at." - Roblimo's last
Defender is Anti-Spyware only. Security Essentials is Anti-Virus & Anti-Spyware combined, so it effectively replaces Defender outright.
really all they are trying to do is fix their broken OS.
It isn't one OS.
Every OS is "broken" in the sense that there are always avenues of attack.
It can't be otherwise so long as mere humans have the final say on which programs can be installed and which programs can be run.
To call something "Malware" is fundamentally a value judgment.
I think the geek would be the first to howl if he could only install the apps approved and certified-safe by Redmond, Cupertino, or his favorite Linux distro.
Its a social problem, not a technical one.
All UAC did was train people to press "Yes" on every dialog even more zealously than before. A system based around asking important questions to people who neither understand or care is not a good one. sudo works because everyone using it does care.
What MS is doing here is clever IMO. Instead of trusting the user to not do anything dumb, they've instead given them a big "Press me to fix your computer" button to wail on. People will see their computer is being slow, hit the button and hopefully the problem will be fixed.
I should be able to submit bug reports directly to MS
secure@microsoft.com
I have actually submitted information about a vulnerability that way. It was fixed.
Viruses don't target server environments. That's pretty much by definition - as they require active user interaction to spread.
Exploits are a different thing, but, really, have you looked at vulnerability stats in Apache vs IIS6/7 lately? Try it, you might be surprised.
It has had much more patches over the year and I'm not conviced that an actively developped open source project like this is more vulnerable because more holes are found. That might be a sign that it's easier to locate in apache than IIS but does not mean they are non existent and unknown to the wrong people in IIS.
To sum it up: you don't care about any numbers I might show you, you just firmly believe that Apache is more secure, simply because it's OSS.
So, is there any point to discuss it further, then?