| ActiveMac | Anonymous | Create a User | Reviews | News | Forums | Advertise |
|
|
|
User Controls |
|
New User |
|
Login |
|
Edit/View My Profile |
|
|
|
|
|
|
|
ActiveMac |
|
Articles |
|
Forums |
|
Links |
|
News |
|
News Search |
|
Reviews |
|
|
|
|
|
|
|
News Centers |
|
Windows/Microsoft |
|
DVD |
|
ActiveHardware |
|
Xbox |
|
MaINTosh |
|
News Search |
|
|
|
|
|
|
|
ANet Chats |
|
The Lobby |
|
Special Events Room |
|
Developer's Lounge |
|
XBox Chat |
|
|
|
|
|
|
|
FAQ's |
|
Windows 98/98 SE |
|
Windows 2000 |
|
Windows Me |
|
Windows "Whistler" XP |
|
Windows CE |
|
Internet Explorer 6 |
|
Internet Explorer 5 |
|
Xbox |
|
DirectX |
|
DVD's |
|
|
|
|
|
|
|
TopTechTips |
|
Registry Tips |
|
Windows 95/98 |
|
Windows 2000 |
|
Internet Explorer 4 |
|
Internet Explorer 5 |
|
Windows NT Tips |
|
Program Tips |
|
Easter Eggs |
|
Hardware |
|
DVD |
|
|
|
|
|
|
|
Latest Reviews |
|
Applications |
|
Microsoft Windows XP Professional |
|
Norton SystemWorks 2002 |
|
|
|
Hardware |
|
Intel Personal Audio Player 3000 |
|
Microsoft Wireless IntelliMouse Explorer |
|
|
|
|
|
|
|
Site News/Info |
|
About This Site |
|
Affiliates |
|
ANet Forums |
|
Contact Us |
|
Default Home Page |
|
Link To Us |
|
Links |
|
Member Pages |
|
Site Search |
|
Awards |
|
|
|
|
|
|
|
Credits |
 |
Most Insecure OS? Yep, It's Linux | |
|
||
| Time: 13:00 EST/18:00 GMT | News Source: WinInformant | Posted By: Byron Hinson | ||
|
According to a new report by the Aberdeen Group, open source solution Linux has surpassed Windows as the most vulnerable operating system, contrary to the high-profile press Microsoft gets for its security woes. Furthermore, the Aberdeen Group reports that over 50 percent of all security advisories issued by the Computer Emergency Response Team (Cert) in the first ten months of 2002 were for Linux and other open source software solutions, which muddles the argument that proprietary software such as Windows is inherently less secure than open solutions. And here's another kicker to the status quo: Proprietary UNIX solutions were responsible for just as many security advisories as Linux in the same time period. Could Windows actually be the most secure mainstream OS available today? | ||
|
Write Comment
Return to News |
Displaying 1 through 25 of 458 Last | Next |
The time now is
4:26:00 PM
ET.
Any comment problems? E-mail us |
| #1 By 20 (24.243.41.64) at 11/26/2002 2:47:49 PM |
|
Ok, before I read the NewsForge BS, let me guess as to what arguments they'll use and I'll see how close I came.
- "That includes all Linux distros! That's not fair as the same bug could exist in 7 distros and they count that as 7 bugs!" - "Yeah, but MS has a longer history of security vulnerabilities!" - "M$ SUX0RS!" - "Yeah, but it's open source, so you can just fix all the problems yourself" - "M$ PAID OFF ABERDEEN!" - "Who sponsored this study?" - "Yeah, but the bugs were all fixed within 17.9 minutes of the disclosure as opposed to 71.9 months for Microsoft." Now, what's my score? |
| #2 By 20 (24.243.41.64) at 11/26/2002 2:59:01 PM |
|
Ok, let's go through arguments, shall we?
"And yet, here I sit with my virus-free, trojan-free Linux box, receiving tons of viruses and trojans from Windows users (that don't affect me), watching news item after news item about sites run on Windows servers getting defaced and broken into. " First logical fallacy: I don't have a virus or trojan for Linux, so therefor viruses and trojans for Linux do not exist. I've never had Polio, so Polio doesn't exist either. He's seen lots of Windows viruses, so therefore it is ONLY because Windows is unsecure that it has viruses, right? What about the other issues that he's conveniently leaving out? Such as the fact that there are many more Windows boxes which makes them a more logical target for virus writers? Second (not even logical) fallacy: That only Windows servers get defaced and broken into. On the average month, Windows and Linux are neck in neck for defacements according to some pretty Linux-biased sources I've seen, so the number is even probably worse for Linux. "According to what I've heard from my many sysadmin and network security specialist friends, no OS or network-connected software is secure unless it's administered properly and security patches are applied as soon as they are available. " The implied follow-up statement is, "Except for Windows". "And then, after I started writing this story, a ZDNet article with the headline Linux utility site hacked, infected came across my monitor, and I started wondering, 'What if these Aberdeen people are right? What if this isn't just Microsoft-sponsored nonsense?'" Ah yes, here's the "M$ funded it, so it must be inherently evil". I'm not sure if MS even funded it, but the arguments to this retarded statement are obvious, so I won't even bother. "A look at CERT's 2002 Advisories and Incident Notes pages was not overly reassuring. Yes, I saw some Microsoft vulnerabilities there that Aberdeen apparently missed, and one for Oracle. " Ah, Aberdeen missed them and somehow he found them. Sure. Uh huh, right. (implication that Aberdeen is incompetent and therefore this report means nothing) "I also think we have enough Microsoft viruses left over from last year that we don't need any new ones this year. " Another implication that only Windows has viruses. Note that his statements are contradictory. Most of the viruses for Windows (like Code Red or Nimda) take advantage of vulnerabilities that have patches that have been out for months. The code red patch was out 3 months before Code Red first showed up. Yet, he claims that properly secured and administered OSes don't have problems. Indeed he's right, but then why all these implications that Windows == viruses? And now for the conciliatory summary paragraph where he tries to wave the wand and make you forget all his assinine and completely false bias: "But the real issue is that we all need to be more security-conscious." Except for Windows which has no security, right? Isn't that what you were saying? "The Aberdeen report points out that the system with the most reported vulnerabilities can change from year to year, but that the overall vulnerability and incident trend is up." Move along, nothing to see here. Linux? What's Linux? "Way up. In other words, whatever operating systems we use, we all need to watch out more for security flaws than we have in the past, and work harder to protect ourselves from them." In other words, Linux ain't that bad, see? Windows and M$ ARE EVAL! EVAL I TELL YOU!! Just ignore this report and put your head back in the hole and you'll be alright! |
| #3 By 20 (24.243.41.64) at 11/26/2002 3:01:03 PM |
|
bas (#5) those comments are worthless.
They're basically "Mozilla is not Linux!". Well, if Aberdeen and the Linux zealots count IE and Outlook Express as Windows, why doesn't Mozilla or KDE or Gnome count as Linux? You guys are so full of logical fallacies and double standards, it's amusing for me to watch you continually embarass yourself. |
| #4 By 20 (24.243.41.64) at 11/26/2002 3:17:20 PM |
|
#10:
of course there are viruses and trojans for GNU/Linux You're one of the first Penguinistas I've ever heard admit that. Most deny even the possibility that one could even exists. I agree about your administrator comments, that's a bad practice. MS has been steadily reducing the need for that. Windows XP goes a long way in that regards. Also, I noticed that you compared XP and Windows 9x as if they're the same. You do realize, don't you, that Windows XP is built on the NT/2K engine and that Win9x is the old 32-bit-hack-on-top-of-DOS crap, right? case closed, you lose Don't be an ass. Especially when there is no case, and you certainly didn't win anything. |
| #5 By 20 (24.243.41.64) at 11/26/2002 3:19:30 PM |
|
Wait a minute, a user can't run KDE AND GNOME at the same time. a user will either be using/have installed GNOME OR KDE.
That's not what I said. Please pay attention, I don't want to have to explain basic things to you. Gnome and KDE both have vulnerabilities, but the NF poster was implying that they shouldn't be counted as Linux vulnerabilities. I argued that if IE and OE vulns are counted as Windows, then KDE and Gnome should be counted too. Please pay attention, it's not that hard to understand, really. case closed, you lose. It's all the more amusing when an idiot becomes and arrogant idiot, as if driving home the point he is, in fact, with no doubt, and idiot. |
| #6 By 20 (24.243.41.64) at 11/26/2002 3:23:36 PM |
|
obviously, pro-MS trolls like daz && eagle && mooresa && baarod will argue that this report is true and flawless.
Where did I say that? In fact, I have not commented at all on the report, only the Penguinista reaction. I have not read the report and cannot comment on it. However, from what I've heard, regardless of the conclusions stated therein, the facts remain the same. Linux vulnerabilities are up, Windows are down. Linux has just as many, if not more problems than Windows. It doesn't take a rocket scientist to figure it out, but it takes a Penguinista like you to deny, deny, deny and stick your head back in the sand. i hope BobSmith or Sodablue actually look at the facts (as they always claim they do) and make up their own minds up. Ah, kissing up a little I see? I'm not going to pay the money for the report and I would imagine they won't either. But like I said above, it doesn't take a rocket scientist to figure out that Linux vulnerabilities are on the rise and that the claims of Linux being more secure are patently false and always have been. case closed, you lose. Third time's a charm eh? I think I get the logic now. If you assert idiotic notions 3 times in a row, somehow they become less idiotic? Isn't that also a logical fallacy? Simply by saying something often it eventually becomes true? -d |
| #7 By 6859 (204.71.100.217) at 11/26/2002 3:49:25 PM |
|
Bas (#10) blathes: "... is that a win32 virus can do more damage if it was run on Windows XP or LindowsOS than a Linux virus can do on a Red Hat/SuSE/Mandrake Linux machine because be default normal users do not have administrative privileges that normal users on WindowsXP or Windows 9x do. "
Keep smoking that crack, baby. A default install of Win2K or XP creates the Administrator (Linux/UNIX equiv of "root") first, then you create a user for the machine. Same under Linux. How do I know that? I use both, moron. Slackware install: after the machine boots up the first time you log on as ROOT, then create the users. Win2K/XP install: after the machine boots up the first time you log on as ADMINISTRATOR, then create the users. How the heck is that any different? It's all about knowing what you're doing. And since you clearly do not. I have to say it is you, bas, who loses. |
| #8 By 20 (24.243.41.64) at 11/26/2002 4:19:07 PM |
|
#23: Parker, trying to squeeze blood from a turnip again? Haven't you learned! :)
It's obvious bas doesn't have a firm grip on common sense. Also, I find it wholely amusing that he used one of my canned Linux responses in my first post for this story. And, as if to add insult to idiocy, he closes his messages with "case closed, you lose". I love it! Keep 'em coming bas, I need more good laughs! |
| #9 By 20 (24.243.41.64) at 11/26/2002 4:24:36 PM |
|
For the record, in Windows XP when you add a user, you are presented with three options for privileges:
[ ] Standard User (Power Users group) [ ] Restricted User (Users group) [ ] Other [ drop down list of all groups ] It doesn't default to Administrators group, it defaults to Power Users. |
| #10 By 20 (24.243.41.64) at 11/26/2002 4:44:20 PM |
|
#27: Do you think ANYTHING will EVER hinder their MS bashing? I think not.
If the problem were facts, logic, and reason, they would've shut up long ago. Throwing more facts at them isn't going to change their mind. |
| #11 By 6859 (204.71.100.217) at 11/26/2002 5:46:46 PM |
|
Dang, I go away for a while and #26 (daz) proves my point at the same time as pimp-slapping the simpleton bas! Sweet!
#35, RoguePenguin, why the heck would you waste time doing that? That won't secure anything, and most likely would only cause irreparable damage to the OS. If the hardware was damaged, then it was broken to begin with, and you shouldn't blame Windows, rather blame the crud you have to work with. You mention "Linux was just installed with all defaults. " Yet you mangle Windows. Gee, that's a fair fight. Ok, how about you go ahead and start renaming random files in the /bin directory. Let's see how long your Linux is up and running. (Which is effectively what you were doing on the Windows side.) This post was edited by Cthulhu on Tuesday, November 26, 2002 at 17:47. |
| #12 By 2332 (65.221.182.3) at 11/26/2002 6:05:25 PM |
|
bas - "If MS put out their source code, then total security advisories issued for that year will break the 1,000 barrier, no doubt."
Hmm... so let me get this straight: 1.) Open source is more secure because there are more eyes and therefore less exploitable bugs because a patch is issued quickly when one is found. 2.) Microsoft code is insecure because their code is close, and therefore the only people looking at it are a handful (less than 5000, for example) of people who might hide the fault, or take a long time to fix it. 3.) If Microsoft opened their code, they would have more than 1,000 exploitable bugs because so many people would go through their code looking for them. Hmm... looks like there is a big contradiction in your logic. Since the only "bad" security bug is one that is found, and therefore exploitable, it would seem that if you want a safe piece of software you do the best you can at trying to find all the bugs before you release it, then you make it really difficult for 3rd parties to find bugs after you release it. Yet you're saying that open source is always more secure... which would obviously not be the case if Microsoft opened their source, as you suggested. Which one is it? :-) |
| #13 By 2332 (65.221.182.3) at 11/26/2002 6:09:48 PM |
|
#45 - How are Windows Admin acounts crippled? There are many levels of administrator, but a domain level admin can do anything and everything, just like a Linux/Unix root account.
The fact is, Windows security is FAR more flexable than Linux security will ever be (without a complete rewrite). Windows allows a hierarchy of security privledges based on group membership, intersections and unions of group privledges, overrides, etc. These things, when used correctly, make for a rock solid security model. Problems arise when either the administrators are stupid, or when there is buggy code involved. The basic security concepts behind Windows are FAR better in every way than those used in the Unix world because NT was designed from the ground up with security in mind, whereas Unix had security slapped on to it years after it was originally created. |
| #14 By 6859 (12.219.44.45) at 11/26/2002 7:17:05 PM |
|
$47, RouguePenguin, writes: "Athlon XP+ 1600 @2100 o/ced".
There's your problem, dingbat, the overclocked CPU. Ever wonder why nobody in the software biz actually cares if their software doesn't work on an o/c system? Hmm? Guess you never thought of that. *YOU* are to blame, not the software or the hardware. Now the truth comes out. Nobody in their right mind with stability on their agenda o/c's their stuff. And as for your 200GB drive, it's probably not fubar, but just needed a low level format, hope you didn't throw it out. Nice going dingus. |
| #15 By 2332 (65.221.182.3) at 11/26/2002 7:55:16 PM |
|
#50 - gg - "NT designed from the ground up with security in mind? Uh :) The real world doesn't seem to agree with you. "
As I said, NT is only made insecure by BUGS, not design flaws. Security problems in NT are typically the result of a poor IMPLEMENTATION of a sound design. This is obviously a problem, and one that I think is being addressed. "most of the security used on NT is pretty much a reiteration of very successful concepts that were used on UNIX." I completely disagree. The way security is handled in NT has only one precursor, and that's VMS. NT has full fledged ACLs, HALs, CAS, etc. There is no Unix in existence that has any of these, so how could they be "reiteration of very successful concepts that were used on UNIX"? For instance, there is only ONE way to secure a resource in Unix and all its variants, and that's through the file system. That's it. That's one reason why all other resources that need to be secured in Unix have a file system entry. /dev/[whatever]... not to mention the fact there is no group inheritance. You have the option of Owner, Group, and User for a resource. Have fun. NT, on the other hand, can apply security to nearly any resource using ACLs, and have group permissions inherit infinitly. With the addition of Active Directory, you have the ability to apply security uniformly and consistently throughout an entire organization to ANY resource. Combine all this with Code Access Security in the .NET Framework, and it is obvious to even the most hardened Unix bigot that NT has a much more robust and flexible security model. Here is an interesting comparison of NT versus Unix security: http://www.ce.chalmers.se/staff/sax/nt-vs-unix.pdf Their conclusion is: "This paper demonstrates that the security mechanisms of Windows NT are slightly better than those of UNIX. Despite this fact the two systems display a similar set of vulnerabilities. This implies that Windows NT has the theoretical capacity of being more secure than “standard” Unix". That conclusion was made in 1998, before Active Directory, the extensive policy system, and Code Access Security were added to the Windows Security arsenal. One has to wonder what their conclusions would be today, since the Unix world hasn't had a single fundamental security innovation in over a decade. |
| #16 By 6859 (12.219.44.45) at 11/26/2002 7:58:51 PM |
|
RoguePenguin, asks “As for the hdd, it runs excellent, where did I say it was bad?” you must not have a good memory, let me quote you, ok?
From your post #31, “The 2 windows pc s are in a garbage dump and have been replaced at MY expense.” Then on your post #47: “The "crud" is an Athlon XP+ 1600 @2100 o/ced EPoX 8kha+ Mushkin level 2 High Performance ddr 2100 I even have the newest maxtor 200 GB hdd sent from soul korea (2 weeks it was available here in the U.S.) sent to me FREE by Maxtor Radeon 8500 plextor cdrw 48x12x48” Oh, and as for what I know about hardware?... well, I know far more than you can possibly understand, what’s more, I also (apparently) understand the idea of logic and the English language far better as well. I’m outta here. |
| #17 By 2332 (65.221.182.3) at 11/26/2002 8:03:17 PM |
|
#51 - bas - "EXACTLY THE OPPOSITE!!! if that is so, why did Bill Gates send out that memo to every Microsoft programmer ordering all developement to stop on all Microsoft products and to do a full security audit of its products/code this year?!!?! "
I usually don't resort to personal attacks, but are you retarded? I specifically said: "Problems arise when either the administrators are stupid, or when there is buggy code involved." A code review is EXACTLY THAT. A CODE review. Not a design review! They're looking for bugs in their IMPLEMENTATION of sound security principles. "obviously you were just joking or you really believe it (in which case you should get your head examined?) " Perhaps you should learn to read and understand logic, and while you're doing that, I'll get my head examined. You need to understand that design and execution are two different beasts. Now, there are a few design/strategy issues that Microsoft has made a mistake on. Here they are in the order they annoy me: 1.) Shipping products in a "all features on" mode. (Fixed in all upcoming releases.) 2.) The occasional default blank password. (Fixed in all upcoming releases.) 3.) Window messages (WMCLOSE, WMSETTEXT, etc.) not having a SID or perm set associated with them. (Fixed when using .NET apps, otherwise good coding practices prevent exploits.) Really, only #3 is a "design flaw". |
| #18 By 2332 (65.221.182.3) at 11/26/2002 8:19:49 PM |
|
#64 - gg - "#60: And I already pointed out that you can add those capabilities to UNIX"
How do you add CAS to Unix? What magical product does that, other that open source versions of the .NET Framework? (Ironically...) "But there are ways to add on ACL's and so on and so on onto UNIX - already existing ones I should mention, not just theoretical (LIDS is an example, and I'll leave it at that) " Does LIDS allow group inheritance of policies and permissions? (Maybe it does, but from what I've read, it doesn't.) At any rate, why are these things not included directly in the kernel. Obviously somebody feels they're needed. "I don't think NT's security is any better: If the implementation is bad, then this is also a flaw in security. Security isn't JUST your model." My assertion has been from the very beginning that the NT security MODEL is far better and more robust than the Unix security model. I agree you need a good implementation, but the implementation is completely seperate from the design, and is an entirely different argument. (A good one, though.) "Standard NT isn't much better than standard UNIX if breaches occur though. And it still isn't designed for security from the ground up." Agreed, if somebody manages to break in, Unix and NT both have issues. :-) But it WAS designed from the ground up with security in mind. That's why it has had all these things (aside from CAS) since day 1. They're not add on products, they're part of the OS. "As for unix handling things through files ... I find that to be a very nice way to access various devices, and I will leave this at that." That's fine, but I pointed that out because it reveals the "slapping on security" methodology that is used in the Unix world. "And for Unix having no innovations? UHm ... they probably have code that does very similar things that NT does." The only security innovation in the past 10 years that I can think of off the top of my head is CAS. The Unix world has nothing like it, unless they got it from Microsoft. (Thanks to the fact that the .NET Framework is an ECMA and soon to be an ISO standard.) "It's no innovation, it's something that was always around and not used (and as the saying goes, if it blaots the kernel, don't put it in, or something)" Not used? What security feature in NT isn't used? All the things I mentioned are certainly used. "I don't care to discuss why it isn't in there - I could only guess at the resoning behind the decision not to include such measures." Ok, but that's besides the point. They don't have it, it's not part of their model, and we're dicussing which security model is better. "Either way. If they're not really used they're not very useful now, are they? How many people do you see using'em on NT?" Lots of people, including me. The most recent security feature, CAS, is the very foundation of a product I'm working on for my company. |
| #19 By 2332 (65.221.182.3) at 11/26/2002 8:37:18 PM |
|
gg - Hmm... this is getting interesting now. What do you mean "mathematically prove your OS' security"? Do you have links on the subject?
"No, I don't think LInux (at least) has CAS" I assure you, nobody has CAS unless they're using the .NET Framework. (Which is available from MS for Windows, FreeBSD, and MacOS 10.2, from go-mono.org for Linux, and from DotGNU.org for Linux... but all non-MS versions go off the Microsoft's design.) |
| #20 By 1845 (12.254.254.105) at 11/26/2002 9:20:11 PM |
|
bas, in answer to the making up my own mind thing...
In July I did a comparison of about 8 distros by version and compared to Windows by version and I came up with similar numbers as the Aberdeen group. I've not done that since August, so the numbers may have changed since then. My thoughts on the matter - the many UNIXes have a hefty amount of vulnerabilities and incompatibilities. The host of Linuxes have many vulnerabilities, but seem to have less incompatibilities with each other as the UNIXes. The many Windowses have many vulnerabilities but are generally pretty compatible with each other. In short, UNIX, Linux and Windows (9x or NT) have many vulnerabilities. You need to choose the OS which best suits your needs. Windows suits all of my needs and so I stick with it. Other people have different needs and make different choices. That's my take on the debate. |
| #21 By 2332 (65.221.182.3) at 11/26/2002 10:08:48 PM |
|
#69 - "everyone's design is based off of someone else's ;) "
Well, that's an impossible statement. Many things are based off other things, but not all, otherwise we would have nothing. Sometimes people have an original idea, and in this case, Code Access Security is an orignal security innovation on the part of Microsoft. "it into an equation in some way and PROVE that you cannot break in." Well, there will ALWAYS be qualifiers. An equation is made up of variables and operators. You might be able to ensure the results of an operation (via an operator), as NT does, but you can never know the variables values ahead of time, as then it's not an equation, it's just a results set. In other words, it's impossible to absolutely prove an OS is secure. We can make assertions like "as far as we know, the OS is secure"... but that's tentative. There is always the possibility it will fail because an unknown set of values for variables is used. "in turn, your OS MUST be the security system, rather than the security system being a feature thereof" I don't really understand what you mean. An OS has functions other than security, otherwise all it would be is a secure system, and pretty much useless on its own. For instance, and OS must manage memory. We can apply security restrictions to that memory management, but that application is secondary to the primary function, which is to manage memory. To design an OS with security in mind from the ground up simply means you determine what resources need to be protected, and you give the OS the ability to protect them. The original Unix protected *nothing* at first, and security for resources (file system) was added later. Everything currently protected in NT had that protection built into the OS from day 1. Memory, handles, filesystem, devices, etc... all have "entries" in the kernel (not really, of course) as "stuff that needs protection". It was therefore designed from the ground up with security in mind. |
| #22 By 135 (209.180.28.6) at 11/26/2002 11:14:11 PM |
|
I just did a Windows XP install this afternoon... Booted up, added to the domain. It didn't prompt me to create any users, and the only users who were listed as Admin on the box were the ones I wanted(Domain Admins and our XP Desktop Admin group).
I think bas has been smoking a Rogue Pengiun. |
| #23 By 135 (208.50.206.187) at 11/26/2002 11:44:14 PM |
|
gg - Yeah, Unix of the 70's didn't include TCP/IP... Today it does.
But what's amazing is how similar Unix of today is with Unix of 1992. There's a lot of better ways to do things out there, but they've never coalesced into a standard way of doing things. Instead they are still add-on products. Why? IBM released AFS to open source, and yet I still don't see it being promoted as a solution with Linux. Why? The implementaiton that I was using back in '93 was incredibly better than NFS, and allowed for ACLs and such. It integrated in with kerberos authentication, it provides location independence like DFS from Microsoft. So why isn't it more common? |
| #24 By 2332 (65.221.182.3) at 11/27/2002 12:42:59 AM |
|
#72 - I'll certainly read up on EROS, but I'm fairly sure what you're talking about is logically impossible without absolute truths.
Since the only absolute truth is truth by definition (1 is a number, because I define it as such), the only perfectly secure OS is one that exists only by definition, and not in real life. Can you make an OS really, really, really, really secure? Yup. Will it ever be absolutely secure? Nope. For instance, for an OS to be absolutely secure, it would need perfect fault tolerance. This, in and of itself, is impossible. Why? Because electrons (and everything else) are governed by quantum physics. The Quantum world has true randomness... it's not just that we can't predict an outcome because we don't have the variables, it's that there are NO variables. So any system designed to be perfectly secure must deal with the quantum fluxations (although the chances of them arising are EXTREMELY small... like it might happen once or twice in 10 billion years). The problem with this is that the fault checking systems are subject to those same fluxations. Obviously, this is a very unlikely situation, but it's just one of many "variables" that we cannot possible predict, and therefore we cannot make a perfectly secure system. One might interject by saying "But RMD, isn't math (1 + 1 = 2) subject to those same fluxations?". Well, no, because math works because we define the rules by which it plays. 1 + 1 = 2 because we define it as so. That equation has no physical manifestation, and therefore is not subject to quantum affects. So, you can indeed create a perfectly secure OS design, as long as you never implement that design. :-) Anyway, it's bed time for me. |
| #25 By 2332 (65.221.182.3) at 11/27/2002 12:31:52 PM |
|
haha...
PHP vs ASP.NET is no contest. (PHP was about on par in most cases with ASP, and ASP.NET is anywhere from 2x to 200x the performance of ASP on identical hardware doing identical things.) At any rate bas, your entire argument was based on the assertion that Aberdeen counted the same bug multiple times, which doesn't appear to be the case. |
|
Write Comment
Return to News |
Displaying 1 through 25 of 458 Last | Next |
The time now is
4:26:00 PM
ET.
Any comment problems? E-mail us |
