- Site menu:
- Latest Discussions
- Latest Newswire
- Archives
- Submit Story
- Story HOWTO
- Syndicate
- E-mail Alerts
Windows beats Linux on security - Wipro Survey
|
|
|
Comments (18)
mmkay. in other news, black is white
They probably don't see many people around them with truly straight faces, and thus must think their faces are the norm.
In other words, they live in a dream world where indeed black is white, night is day, and pigs have their own airline ;) .
I believe large impositions of money can help feed those hallucinations...
In other words, they live in a dream world where indeed black is white, night is day, and pigs have their own airline ;) .
I believe large impositions of money can help feed those hallucinations...
Serious flaws in Wipro study
To quickly point out six flaws in this study:
1) Previous research shows most of the companies don't have any idea of TCO-difference between Linux and Windows, so how would they now cost-differences when it comes to patching?
2) quote] Independence Air... with Linux they had to rely on consultants to tell them if their system was secure or not, while in the case of Windows, they could depend on Microsoft to inform them and provide the necessary updates. [/quote] Right. They just have incompetent Linux administrators if they need a consultant for Linux, and they shouldn't trust MS when security is an issue. These people are really stupid, or paid by MS.
3)
4) Next remark, let's say you run a Gentoo system (I know Gentoo, but the same is true for almost every other distro as well), with Apache, MySQL and Perl/PHP for example, and compare it to some Windows server which needs some 3d party software as well, but doing the same, and you're going to update; Using some Linux distro, you could update all 3d party software without any more effort than updating your base-components. (something like Emerge -pu world) When using Windows, you'd have to check all 3d party software seperately, and update it seperately, using different procedures.
5) I read 'MS; Monthly rollout of updates'. So when it's the 7th day of the month and there's a critical bug, you'd have to wait 21 days?
6) MS milestone releases (for example Windows 2000 ==> XP, XP_SP1==> SP_2) make it more difficult to update than some more gliding scale.
Concluding, I don't take this study too seriously, but still I'm afraid others will.
1) Previous research shows most of the companies don't have any idea of TCO-difference between Linux and Windows, so how would they now cost-differences when it comes to patching?
2) quote] Independence Air... with Linux they had to rely on consultants to tell them if their system was secure or not, while in the case of Windows, they could depend on Microsoft to inform them and provide the necessary updates. [/quote] Right. They just have incompetent Linux administrators if they need a consultant for Linux, and they shouldn't trust MS when security is an issue. These people are really stupid, or paid by MS.
3)
Quoting: Survey respondents consistently overestimated the number of Windows vulnerabilities, while underestimating those for OSS.I think they underestimate the vulnerabilities of Windows as well, because of the biased vulnerability-rating MS uses.
4) Next remark, let's say you run a Gentoo system (I know Gentoo, but the same is true for almost every other distro as well), with Apache, MySQL and Perl/PHP for example, and compare it to some Windows server which needs some 3d party software as well, but doing the same, and you're going to update; Using some Linux distro, you could update all 3d party software without any more effort than updating your base-components. (something like Emerge -pu world) When using Windows, you'd have to check all 3d party software seperately, and update it seperately, using different procedures.
5) I read 'MS; Monthly rollout of updates'. So when it's the 7th day of the month and there's a critical bug, you'd have to wait 21 days?
6) MS milestone releases (for example Windows 2000 ==> XP, XP_SP1==> SP_2) make it more difficult to update than some more gliding scale.
Concluding, I don't take this study too seriously, but still I'm afraid others will.
It was commissioned by Microsoft, so it has to be true.
Rely on Microsoft?
"Independence Air..... found that with Linux they had to rely on consultants to tell them if their system was secure or not, while in the case of Windows, they could depend on Microsoft to inform them and provide the necessary updates."
Yeah, that will work. Remind me never to fly on Independence Air!
Yeah, that will work. Remind me never to fly on Independence Air!
Hmmm.
Or in other words, Linux consultants told us stuff we didn't understand and claimed this Linux was secure, but, since we know nothing about IT, we couldn't be sure. Whereas with Windows we know it will never be secure so we don't need to pay consultants to tell us stuff we don't understand.
And when something goes wrong we can shrug and say "bloody Windows" and it's Microsofts' fault.
salparadise: ever notice how much your typical windows admin knows about the theoretical underpinnings of his O/S?
This type of "wisdom" presented here plays right into their mind-set.
This type of "wisdom" presented here plays right into their mind-set.
cronned "urpmi auto-select" vs hotfix bingo?
Hmm. Let me think about this. Which is easier?
"apt-get update install"? Oh, the pain! Three whole words! It'll never fly!
Biassed? Deceptive? Misleading? Atypical? A study from Microsoft? Never!
"apt-get update install"? Oh, the pain! Three whole words! It'll never fly!
Biassed? Deceptive? Misleading? Atypical? A study from Microsoft? Never!
apt-get update install
Don't tell anyone, but if you put that in cron, then admin costs of updating a machine approach zero. The only cost involved would be if somehow an upgrade failed and it required some attention. With a local apt repository, you can fine-tune what gets updated, too.
Don't tell anyone, but if you put that in cron, then admin costs of updating a machine approach zero. The only cost involved would be if somehow an upgrade failed and it required some attention. With a local apt repository, you can fine-tune what gets updated, too.
chris: I had a good debian friend tell me never to put this in cron (it was over 3 years ago though -- has this warning expired?).
--FeriCyde
--FeriCyde
Methinks there is a typo here- 'apt-get update install' is not a valid command. I put
apt-get update && apt-get upgrade
in a weekly cron job, spose I could do daily but meh, why bother. Also you want it configured so that dpkg --configure doesn't ask you for silly decisions like locales and time zones.
If you're using only official Debian repositories any risk is tiny, in fact compared to applying winduz patches it's infinity plus one safer. Never ever ever apply a windoze patch without extensive testing, and even then there is no way to test for every possible bad thing that can happen, and believe me there are a lot of them. You may have a driver conflict, or it may overwrite some system files and bork something else, or an application might make it choke, or it comes spyware-ready, or some weirdo windunce networking protocol goes insane, or the Registry becomes possessed, or sheeesh I'm getting flashbacks just thinking about it, and they ain't the fun kind.
Every linux system I've ever had control of in the last 2-3 years has been set up to download automatic updates. The worst thing that's ever happened is the occasional packaging conflict with up2date or yum. Be at peace little Tuxers- keep your systems updated and relax. It's winschmuck admins who have to lay awake at night having nervous sweats over these issues.
apt-get update && apt-get upgrade
in a weekly cron job, spose I could do daily but meh, why bother. Also you want it configured so that dpkg --configure doesn't ask you for silly decisions like locales and time zones.
If you're using only official Debian repositories any risk is tiny, in fact compared to applying winduz patches it's infinity plus one safer. Never ever ever apply a windoze patch without extensive testing, and even then there is no way to test for every possible bad thing that can happen, and believe me there are a lot of them. You may have a driver conflict, or it may overwrite some system files and bork something else, or an application might make it choke, or it comes spyware-ready, or some weirdo windunce networking protocol goes insane, or the Registry becomes possessed, or sheeesh I'm getting flashbacks just thinking about it, and they ain't the fun kind.
Every linux system I've ever had control of in the last 2-3 years has been set up to download automatic updates. The worst thing that's ever happened is the occasional packaging conflict with up2date or yum. Be at peace little Tuxers- keep your systems updated and relax. It's winschmuck admins who have to lay awake at night having nervous sweats over these issues.
Quoting:compared to applying winduz patches it's infinity plus one safer.
I think that's an exageration. It's never possible to measure the risk of doing a Windows upgrade...
Paulie, Paulie, Paulie:
You Open Source guys, smugly confident in your ability to assess the risks built into code that can see but rarely bother to read, persist in exaggerating the danger of Windows upgrades.
Truth is, it's far SAFER to automatically upgrade Windows than it is to automatically upgrade Linux or any of the BSDs.
If you talk to many Windows users for very long, you'll hear terrible horror stories of viruses and trojan horses, some merely annoying and some destroying valuable data.
Given how bad off Windows users are, the risk that updates will leave them worse off is pretty minimal. It's the old nothing to lose thing...
You Open Source guys, smugly confident in your ability to assess the risks built into code that can see but rarely bother to read, persist in exaggerating the danger of Windows upgrades.
Truth is, it's far SAFER to automatically upgrade Windows than it is to automatically upgrade Linux or any of the BSDs.
If you talk to many Windows users for very long, you'll hear terrible horror stories of viruses and trojan horses, some merely annoying and some destroying valuable data.
Given how bad off Windows users are, the risk that updates will leave them worse off is pretty minimal. It's the old nothing to lose thing...
Dino, while you make a good point there, and I can see it, the truth is that a good hat would cover it up nicely and then you could go out in public without shame.
Yes, I see the bit about "Why not risk shaking the San Andreas Fault -- can't hurt". Worse though is this quote here:
First, I think it's rude for you Windows communicating types to categorize people in with a group. When you do that, you often make all kinds of stupid assumptions about their belief system, religion, what they're wearing (etc) -- you generalize, in other words -- and you could be wrong about it. For example, I talk to Windows people all the time, so you can wipe that smug grin off of your face and take off that Microsoft-logo'd polo shirt -- it clashes with the plaid pants, after all.
Second, I also think it's rude to swear.
Yes, I see the bit about "Why not risk shaking the San Andreas Fault -- can't hurt". Worse though is this quote here:
Quoting:You Open Source guys, smugly confident in your ability to ass
First, I think it's rude for you Windows communicating types to categorize people in with a group. When you do that, you often make all kinds of stupid assumptions about their belief system, religion, what they're wearing (etc) -- you generalize, in other words -- and you could be wrong about it. For example, I talk to Windows people all the time, so you can wipe that smug grin off of your face and take off that Microsoft-logo'd polo shirt -- it clashes with the plaid pants, after all.
Second, I also think it's rude to swear.
apt-get install cron-apt
for automatic updates done right. It should be useful even on unstable/testing to leave in the default "download-only" mode. In stable you can even configure it to install updates.
for automatic updates done right. It should be useful even on unstable/testing to leave in the default "download-only" mode. In stable you can even configure it to install updates.
Quoting:Second, I also think it's rude to swear.
Hmmm, Paulie. We don't often agree, but you may have a point there. I think I will take a walk down to the dam now, and assess the rash items I may have written.
Are any of you real Debian sysadmins?
Would you really do an apt-get install on anything other than on the stable release? In that case, one is at least assured the changes have been heavily tested and are comprised of necessary security fixes.
Now I know I am ignorant, hence, I am setup under cron-apt do an "apt-get update" that downloads the package changes and summarizes the changes that will occur if an apt-get install is run. Should a non-zero "remove" be present caution would be advised.
I am running a mixed version of testing plus unstable. There is too much I do not know about Debian, other than I am not going back to rpm's ever! With Fedora Core on my laptop, yum on the command line on a rapidly changing small core of packages suffices for me.
Would you really do an apt-get install on anything other than on the stable release? In that case, one is at least assured the changes have been heavily tested and are comprised of necessary security fixes.
Now I know I am ignorant, hence, I am setup under cron-apt do an "apt-get update" that downloads the package changes and summarizes the changes that will occur if an apt-get install is run. Should a non-zero "remove" be present caution would be advised.
I am running a mixed version of testing plus unstable. There is too much I do not know about Debian, other than I am not going back to rpm's ever! With Fedora Core on my laptop, yum on the command line on a rapidly changing small core of packages suffices for me.
Quoting:I think I will take a walk down to the dam now
Crap! There you go again! How rude!
Post a comment
A free account is needed to post. Write your comment here first: your text is kept in this browser while you sign up or log in, and is waiting for you when you come back.
Your message:Markdown: **bold**, *italic*, [text](https://link), - lists, > quotes, $x^2$ for math. Formatting help
Sign up and post (it is free) or Log in and post
How do these authors keep straight faces?