Slashdot Mirror
Fighting Claims That Open Source Is Insecure?
Lumpy asks: "Lately there has been a HUGE push by Certified Microsoft Professionals and their companies to call clients and warn them of the dangers of open source. This week I received calls from 4 different customers that they were warned that they are dangerously insecure because they run Open Source Operating systems or Software because 'anyone can read the code and hack you with ease' they are being told. Other colleagues in the area also have noticed this about 3 Microsoft Partners or so they claim have been going out of their way to strike fear of OSS in companies that respond with 'yes we use Open source or Linux' when the sales call comes in. I know this is simply a sales tactic by these companies that will remain nameless, but how do I fix the damage caused by these sales tactics? I have several customers that now want more than my word about the security of the systems that have worked for them flawlessly for over 5-6 years now with minimal expense outside of upgrades and patching for security. Does anyone have a good plan or sources of reliable information that can be used to inform the customer?"
I think one of the most powerful ways to demonstrate open source is to show people how much they are using without even knowing it.
On a couple of occasions I've spoken to IT people who have said things like "we'd never touch open source because..." and then I've been able to point out multiple ways they use it without realising it. If they use google, if they use email, if they use many websites, then they're using open source software. Many bits of hardware contain open source code (wifi boxes for instance). Many companies are using Apache for their web sites without realising it.
Another good argument is just to spout off a list of Fortune 500 companies who use open source to run their websites. "it's secure enough for IBM, but not secure enough for you?" is the type of argument that's difficult to counter. Very often they just don't know much about it.
The problem you have to fight in people who say things like "open source is insecure" is their ignorance.
Ask your customer a simple question in reply:
Does that fact that closed source software hides it's defects mean that it doesn't have any defects?
Or, how about the really important one:
Would you rather be at the mercy of your vendors to disclose (against their own self-interest) and fix security issues (on their own timetable); or would you rather have a multitude of people, who are dedicated to the values of openness and transparency, constantly striving to keep open source software as secure as possible?
Insisting on "correct" English is like saying that there is only one, definitive recipe for chili.
"'anyone can read the code and hack you with ease'"
Likewise, anyone can read the code and repair it with ease.
High-profile projects run by responsible people will benefit from the "many eyeballs" approach and be better quality than if they were closed-source run by a team of a few or dozens of people.
The FUDsters do have a point when it comes to out-of-date or low-profile software:
If an adversary knows YOU run last-year's version of apache or that you run some obscure open-source database on your web site, they can find and exploit bugs that are either already fixed or that nobody else is looking for.
The moral of the story:
1) Stay current with security patches
2) Hide what you use from the adversary. If they don't know you run ObscureWebServer 1.0, they don't know to try attacking it first. Keep them guessing.
3) Make sure the official vendor/caretaker takes reports of security breaches seriously and is willing to consider patches from the community
above all,
4) Don't depend on your software's security to protect your assets. Make sure you have good backups. Train your employees against social engineering attacks.
Security is but one of many factors that go into the open/closed source decision.
For me, two of the biggest factors are:
1) if the product is abandoned or sunsetted, I can maintain it myself or hire someone to maintain it
2) If I don't care about paid-for support, I can use the product on as many machines as I want without worrying about "product activation" or getting sued.
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
Open Source knows the source is going to be open. So the security model starts from knowing that will be the case.
Closed Source security thinks that no-one else knows what is in there. THINKS being the operative word. Maybe they've worked on that assumption, and just obscured the holes rather than fixed them. Maybe they've left some deliberate backdoors, on the grounds that no-one else knows they are there. Possibly not, but you don't know that.
The MS people are correct to say that it is easier to construct an exploit for one category of security hole if you've got the source. But that means that those sorts of holes don't get built in the first place.
"I Know You Are But What Am I?"
saying that software is 100% bug free, or not exploitable is a complete fallacy.
all software has bugs in it, there is no such thing as a completely secure application.
the point of open source software is the more eyes you have looking at code, the easier it is to find and patch these bugs...
the problem with closed source software is that the bugs aren't easily as found, and certainly not easy to patch, especially since only few have access to the source. So while the bugs exist, they go unfound, generally found first by some obscure hacker who may or may not have the best intentions.
To answer the articles question you have to point out the shortcomings of all programs, and that for ever malicious hacker scanning source code to determine flaws in any given open source project, there will most likely be any number more of benevolent people trying to stop him.
I find what Adobe said about software development for Linux simply being hard more interesting than the security question. My experience has been that most people expect any platform to not be as secure as they'd like the same way they've expected their computers to not be as stable as they'd like. The thing they need is good software and now Adobe is pointing out that writing and maintaining software for Linux is difficult because, despite some good efforts, there still is no standard definition for what a linux system is or contains...
Reality is nothing but a collective hunch.
You can also make an analogy to government using the parent's ideas. Would you rather have an open, transparent government where you can inspect each and every process or would you rather have a closed, secretive government where anything can happen without your knowledge?
Love sees no species.
I have several customers that now want more than my word about the security of the systems that have worked for them flawlessly for over 5-6 years now with minimal expense outside of upgrades and patching for security.
Try IBM,
http://www-1.ibm.com/linux/opensource/
Download some of the report PDFs and send them to your clients.
This week I received calls from 4 different customers that they were warned that they are dangerously insecure because they run Open Source Operating systems or Software because 'anyone can read the code and hack you with ease' they are being told.
I'd have your sales rep call your clients and let them know that your company shares thier concern. At the same time remind them of SQL Slammer, Code Red, Melissa, Blaster, etc. Point out all the other companies using OSS products, Google, Wall Street, etc.
Of course I'm just a programmer, so take my comments with a grain of salt.
Enjoy,
It's just the normal noises in here.
The fountain of knowledge that is Wikipedia has this article, http://en.wikipedia.org/wiki/The_Cathedral_and_the _Bazaar, which is interesting. Its an essay/book about open source development, and there is a link to the full text in the WP article. There's a chapter about why open development is good (from a quick look at te text), and I know I've read similar-minded texts on sites like gnu.org and fsf.org, but was unable to find them. I think Cory Doctorow has written some good articles about secrets and the management of them, but I think his are more DRM musings, though the same principles apply to proprietry software vs. open software.
Articles about why SSH etc. are secure, even though their inner workings are wide open to the world, may be helpful too.
Car analogies break down.
A lot of this centers around that because the source is exposed, anyone could exploit it for flaws.
Consider which is less secure, a project whose source is always available, or a project whose source suddenly becomes available? I would guess that since Microsoft has never officially had its source be in the hands of hackers, there are TONS more exploits there that if you did see the source, you would easily find. Since OSS is always visible, people are quick to point out and fix various holes. This is a much more effective way to manage source control, since any fixed number of people can only read so much into a massive body of source code.
Also, not anyone can modify the actual gold master source for an OSS project, so it's not insecure in that way.
stuff |
One word:
botnets
Then you can explain how it's actually the closed source OS that is the most damaging.
Hell, just show them some apache logs that are still constantly being hit by things like IIS servers still infected with Sasser, years after it should have been eradicated.
...Rob
The American Dream isn't an SUV and a house in the suburbs; it's Don't Tread On Me.
Here's an example with bank vaults. Suppose I have two identical looking bank vaults, one showing the schematic and one hiding it. Which one can you exploit more easily? The vault showing the schematic has nothing to hide... if it's secure, then seeing the schematic doesn't make getting through a foot of steel any easier. However, the one not showing the schematic might have reason not to show it from a security standpoint, i.e. that little screw in the back of the vault, that if you just were to unscrew it, you could break in. Whether the system is strong or weak, open source will expose that. So from a security standpoint, your system's strength doesn't lie in its obscurity, it lies in your ability to disclose exactly what it is doing and still not be compromising its security.
Strong security discloses the facts, i.e. : "here's the pile of money, and there's the guy that will shoot you if you try to take the money."
stuff |
then simply note that that the assumption being made is that all software is flimsy. The point of open source is to subject software to examination so that it is strenghtened.
Here's a good analogy. If I walk into my local bank branch, I can see the bank vault behind the tellers. The massive, foot thick steel door stands wide open, and if you look, you can see the network of gears and lever bars that are needed to for a person of ordinary strength to drive home the dozen massive two inch hardened steel bolts that secure the vault when locked.
Now, the design of the door mechanism might useful information for me if I wanted to break into the vault. The bank is placing this information in full view in part to reassure its customers. But it also deters people like me from even trying. Yes, it reveals potential vulnerablities, but on balance the message to me is that there are more practical ways to make a buck.
Being confident enough to expose your vulerabilities is a good sign, not a bad one.
Hiding vulnerabilities is not a sign of strength. If the customer can't see for himself or through an agent that a piece of software is secure, why bother making it secure? And hiding source code doesn't hide vulnerabilites. A burlgar can make use of floor plans if he has it, but not having floor plans is no deterrant. Furthermore, unlike you, hackers can reverse engineer the source code, so the only party left in the dark is you.
Here's a good question to ask: has the software vendor subjected his product to a responsible and independent third party security audit? Why not? Companies disclose source code all the time under NDA, so there's no risk there. And it isn't expensive in the grand scheme of things, unless they audit reveals the sofware to be so insecure the vendor has to throw a lot of it out.
Post may contain irony: discontinue use if experiencing mood swings, nausea or elevated blood pressure.
Sure, but have you seen how a lot of bugs are being found lately? Fuzzing. You can fuzz both closed and open source software the same way. Sure, if you had the code for it, you could look at Joe Schmoe's web software and look for input validation bugs, and maybe find one after a while. Or, if it was closed source you could fuzz it and find a bunch of vulnerabilities you probably never would have thought of looking for.
My point is, insecure software is always going to be insecure, whether it is open or closed source, and don't let someone kid you into thinking that one has an aboslute, inherent advantage over the other.
There are two types of people in the world: those who divide people into two types and those who don't.
Re: your sig-line: "Only you can prevent hotlinking. Well, you and mod_rewrite."
When read your post, this banner ad graced the top of the page.
Cosmic Coincidence or intelligent ad placement gone haywire?
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
Along with any number of other good answers, I'd also point out that Microsoft has a very poor security track record and is hardly in a position to be making ominous threats about other people's security.
Here's a search for "Microsoft" on the Open Source Vulnerability Database. ("Open Source" here refers to the nature of the database, not covering only open source products.) Pop in any other large closed-source vendor you can think of and you'll find something. ("Oracle" is another personal favorite. It may have "Enterprise-class" performance, which I can't vouch for either way having never used it, but it sure doesn't have "Enterprise-class" security.)
I think the main problem with the implied argument is that you don't need source code to find security vulnerabilities (in fact it might not even be helpful given the other cracking techniques you can use), but you do need it to fix them, with rare exceptions.
"Ask your customer a simple question in reply:
Does that fact that closed source software hides it's defects mean that it doesn't have any defects?"
To attain exactly, what?
Just to follow your argument, here comes the obvious answer to your "counter-question":
Of course closed software has its defects. But then, its defects are hidden, aren't they? So they are obviously more difficult to exploit, and I prefer to have a software its defects are difficult to exploit rather than one which is easy to exploit. I'm questioning my confidence on your ability to have the things done if I have to explain to you such an obvious thing!
"Would you rather be at the mercy of your vendors to disclose (against their own self-interest) and fix security issues (on their own timetable); or would you rather have a multitude of people, who are dedicated to the values of openness and transparency, constantly striving to keep open source software as secure as possible?"
Hummm... at the end of the day, a USA corporation may be held legally liable. Do you really expect me to try to recover damages from a stinky teenager deep in Soviet Russia (where teenagers stink you) that happened to develop some seemingly cute software in his spare time?
No, the answer has been already told. If they really are paying attention at such stupid arguments like those from 'M$ drones', they are ignorant about these issues, and the best course of action is enligth them in such a way they can understand:
Look at IBM: they extensively use open source and it seems they are not going into bankrupcy anytime soon.
Look at Google: they critically use open source, they have an ashtounding computer-base all around the globe and still it doesn't seem like they are hacked everyday, do they?
You can ask a question *then*:
Look at IBM or at Google, or at almost every Fortune 100 out there; they do well using open source. Don't you find suspicious the only ones pesting about open source are companies (Microsoft and its VARs) that *do* would go bankrupcy if open source took the computer world for a raid?
You can also point out that, when bugs are found, they tend to be fixed very rapidly, frequently within hours of their discovery. Since the source code is available to everyone, anyone affected can create an update to fix the problem. This happens exceedingly rarely in the closed-source world, despite the large numbers of bugs encountered.
PHEM - party like it's 1997-2003!
> because 'anyone can read the code and hack you with ease' they are being told
Hm. In the open source arena, if someone is reading your code, they've obtained it legally. Most people who read OSS code do so to improve the code--not specifically for the purpose of creating a full-fledged exploit with it.
In the Windows world, if someone is reading your code then they are either: 1. an employee of Microsoft or 2. someone who stole the code. In the first case they're ethically barred (not supposed to. *ahem*) from using their corporate knowledge to hack you. In the second case they've already established themselves as a criminal.
Which situation makes you feel more comfortable about knowing that other people can read your code? I choose OSS.
the NPG electrode was replaced with carbon blac
Better questions would be:
Where are the articles about companies losing data due to defects in OSS?
Now where are the articles about IE (for example)?
Once they compare them, they will see the light.
Intron: the portion of DNA which expresses nothing useful.
You might say, yes yes, I know about all that, but you can't actually do that in practice. I would bet, though, that some of the early electronic calculators were proven correct. The people making them in the very beginning were probably interested in such things. Perhaps some apps running on MIT LISP machines were also proven (LISP is easiest to prove, and the MIT AI lab people are the type to do it), although in this case it is unlikely that the entire platform up to the app was also proven. So it is not so cut and dried as to allow you to say that there are no completely secure apps. Reasonable, useful apps today, probably none are completely secure, since I doubt that any kernels are completely secure if for no other reason. But nonetheless, it is possible to have 100% bug free, 100% secure software.
SIGSEGV caught, terminating
wait... not that kind of sig.
Yes, and no. You can't make "bug free" software, because one persons feature (or lack of) is another's bug. However, I believe, you can make secure (read: no remote exploits) software. That's a much smaller scope you have to defend against, and it's mostly testable. Also multiple people have done it, or claim to have done it ... including myself.
ustr: Managed string API with ave. 44% overhead over strdup(), for 0-20B