Free as in Beer has caused Heartbleed (and Much More)
Heartbleed is a bit over one month old now. A bug significant enough to have its own Wikipedia page. Today, we’re going to look into how wrong we have been in assuming that Open Source software is more secure than commercial software, because of our thinking that source code is open and that many developers are looking into it.
Free as in Beer
One of the core principles of Open Source software is, well, that it is open and that this openness is free in one way or another. This allows us developers to browse source code of third-party software and libraries for various reasons:
- To learn from it
- To copy it (under the terms of the respective license)
- To modify it (under the terms of the respective license)
- To verify it
Proprietary software often does not have the above attributes in exchange for warranties. When you read through the millions of lines of unintelligible legal Microsoft text, for instance, you will see that Microsoft gives a couple of guarantees, also with respect to security and how security flaws are remedied.
The warranty that is legally shouted at you by OpenSSL is this one:
THIS SOFTWARE IS PROVIDED BY THE OpenSSL PROJECT “AS IS” AND ANY EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE OpenSSL PROJECT OR ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
In fact, you don’t have any warranties. Specifically, there is no liability for direct or indirect loss of data, as has happened because of Heartbleed.
Obviuous, right? Because it’s free as in beer, any consequences resulting from the Heartbleed bug are entirely your own fault, if you’re using OpenSSL, or any other open source software that includes OpenSSL under similar terms.
Yes. You should have known that you were vulnerable, because you actually could verify it. Did you? Of course not. Did we? No way we’re delving through all that C code. Did our “suppliers”? We don’t even know. We’re using WinSCP extensively, and check this out, WinSCP was affected by Heartbleed! The developers over at WinSCP have been nice enough to fix this issue quickly and release a new version. They didn’t have to do this. You know what? Let’s hit that donate button, right now to thank them.
“They” should pay
OK, but again, wasn’t Heartbleed supposed to never even happen? Isn’t this Open Source? Doesn’t anyone (read: “they”. Because, I don’t have time) review such code, especially if it’s that widely used? Isn’t Open Source security much better than closed source security, which is essentially security through obscurity?
Eberhard Wolff, a well-known German freelance consultant and trainer recently replied this to me:
Yes, I’m sorry, Eberhard. But that’s just the case. Security is often neglected in many pieces of software, both commercial and open source. It’s not that the openness helps anyone discover things, security issues are very hard to discover per se. Also in commercial software. Remember GOTO fail? GOTO fail also affected SSL, but only in Apple software.
The issue lies elsewhere, though. Because Microsoft somehow manages to patch all security issues in virtually no time. They fix things so fast that users get frustrated about the sheer frequency of fixes ;-)
But Microsoft has a lot to lose. Pretty much 90% of desktop OS market share, that’s what they’ve got to lose. They’re paying a high amount of insurance money, invested in security teams that are penetration testing their own software to look for leaks. Yes. The vendor themselves are doing this. And yes, it costs money. And yes, you’ve already paid that when you bought Windows (or Office for that matter).
Thanks to that publicity there has been an outpouring of grassroots support from the OpenSSL user community, roughly two hundred donations this past week along with many messages of support and encouragement. Most were for $5 or $10 and, judging from the E-mail addresses and names, were from all around the world. I haven’t finished entering all of them to get an exact total, but all those donations together come to about US$9,000. Even if those donations continue to arrive at the same rate indefinitely (they won’t), and even though every penny of those funds goes directly to OpenSSL team members, it is nowhere near enough to properly sustain the manpower levels needed to support such a complex and critical software product. While OpenSSL does “belong to the people” it is neither realistic nor appropriate to expect that a few hundred, or even a few thousand, individuals provide all the financial support. The ones who should be contributing real resources are the commercial companies and governments who use OpenSSL extensively and take it for granted.
I can feel with Steve. US$9,000 to improve OpenSSL, a software that has added so much value to all of our servers and clients. That’s hardly enough for even 2-3 bug fixes, depending on whose salaries you’re paying.
But Steve tells off the “commercial companies” and “governments” to not have paid enough. But who are these “commercial companies”?
Meanwhile, at Data Geekery
We know similar stories, of course. When we have transitioned from completely Open Source to dual-licensed, we heard a lot of complaints by users that we have never heard of up to that moment. Users who have never donated or contributed code, bug reports, manual improvements or anything. This is OK. We were giving jOOQ away for free under the terms of the ASL 2.0 as part of a extended market analysis. We didn’t expect donations.
Of course, we have also disappointed one or two people who were hoping that jOOQ would remain “free as in freedom”, who have contributed, who have participated, and who now felt deceived. And we’re genuinely sorry about that. But participation doesn’t pay for our bills, money does. So we reached out for money (luckily, we always owned the code as we paid contributors of larger contributions, and had them all sign a CLA, so we could actually do this step, from a legal perspective).
But let’s again focus on the people looking for free beer. No contributions. No feedback. Free rides. And then, when asked for money, they’re pissed (even 8 months later!).
Fair enough. Our price has risen for that particular user. From free to something. They, too, have a right to be frustrated about this change, which they might not have expected. But they, too, have taken “free” for granted without proper reason. Without verifying.
We don’t believe that “THEY” should pay for our licenses. We don’t think that “THEY” should pay for enterprise support. There is no “THEY” in free software, there’s only all of us, because who decides which project is really important enough to deserve some funding from “THEM” and which one isn’t? All attempts of “fairly” distributing money will inevitably lead to corruption, misuse, abuse, and eventually, to a lack of innovation. We’ve known that from other industries.
So, let’s simply establish two facts:
- Free software is lying around in the streets. It is publicly acclaimed to be AS IS software. If we’re using it, it is our own risk. And if it goes wrong, it is our own fault. No one else’s. Not the big companies’ (who are already investing tons of money in Open Source software), and not the governments’ (same thing there). Stop pretending that FOSS is in any way better or less of a legal risk than commercial software. That’s just not true.
- In fact, there is no such thing as free software. “Free” is a price we’re paying as a down payment (or non-payment). The costs will or might arise later. If we’re lucky, the thing will remain “free”. But if we’re professionals, then we’ll insure ourselves against any risk arising from free “AS IS” software. This means that we’ll give back (= “pay”) in another way. Through donations, through bug fixes, through verification.