“We did not exploit any flaws or vulnerabilities. The cameras have a built-in maintenance backdoor, which allows anyone with super admin privileges to access a root shell on any camera of any customer at the click of a button.”
I'd like to add a bit of context to how security cameras most often are installed.
In the industry in general you have producers of the equipment and you have buyers, but in between there you have integrators. The integrators plays a crucial role when installing big systems. They win the bid for an installation and carries out the work. This means that there is seldomly any direct path between camera producer and the customer. For the producer to get access to footage they must go through the integrator, so the friction is non-trivial.
Direct contact producer <=> buyer might happen in the small case, like a store with a single camera or you placing one at home.
My guess (!) is that verkada tries to pry away the integrators with a simpler model for installation.
Most larger producers now have cloud offerings, which could have some similar vulnerabilities to those mentioned in the article. However, my impression is that security is taken VERY seriously. Not just lipservice, but in practice. This makes sense as it is a key selling point and the larger buyers are competent judges of this. This is in stark contrast to the "typical" hacked target, which seems to be autoshops and hospitals (I am generalising to get through a point, I am not sure what the most common victim is).
This really struck me also. I work in the relevant industry (we make cameras etc.) and there is always a bit of pain to get user footage. This is how it should be! To have everything from source code to customer material accessible to an admin is bottom-tier thinking. Why not just rename your "admin" to "GOD" and then ask yourself if you have any single point of failure?!
I do NOT want to sound smug, but there is a little bit of amateur hour going on here both from buyer and seller. High value and large targets (like airports) and more established sellers usually don't work like this, and that's for a reason.
It's not just that it's a single point of failure, it's that as a customer I do not want any admin who is feeling curious to be able to snoop on my footage with a click.
I don't know how "established" this company is, but their customers appear to include city governments, hospitals, and Tesla motors, which I would consider "high value and large targets".
Makes me suspicious of the whole industry. If others in the industry dont' want that, time for some industry codes and audits and self-regulation.
I've added a link to IPVM to the parent to my comment that might interest you!
Regarding established: I might be wrong! I willingly admit that I knew nothing about verkada some days ago. Seems to be relatively new (5 year-ish) and "classic" Silicon Valley in that they push hard for growth to get their valuation up and try to "disrupt" by running everything in the cloud. More sales people than R&D, which I think is uncommon.
Verkada runs full lock-in, so if you buy a camera from them you have to buy their services. This is again relatively uncommon. Most of the industry supports the ONVIF standard, so you can run the hardware you bought with different software solutions. If you want encryption at rest, no problem. You just make an on-premise solution with full encryption. With verkada you can't do that (incidentally verkada have mocked ONVIF due to alleged security concerns, but obviously it undermines their business model with full lock-in).
Since combining verkada and other hardware would require parallel systems I made an educated guess that most customers would be places without previous hardware and/or less concern for the long run. Most large and high value targets have previous hardware, but certainly there are exceptions. And as stated earlier, I might be wrong:)
And lastly, you should be suspicious! Last time I bought a car I was very suspicious. I like the car I did buy very much, but next time I will be just as suspicious again. That's how things should be when it's about trust and high impact.
The line-by-line explanation is very handy, thank you for that. I find that one of the hard balances with using Emacs is whether to fully understand everything you put into your conf-files or just accepting some degree of copy-paste.
Thank you for sharing your feedback. Understanding every line of .emacs was indeed the intention behind the line-by-line explanation.
If one does not wish to understand the customizations completely, then Portacle is a great starting point. However, my intention here was to promote a DIY approach to customizing Emacs into a CL dev environment.
I believe that with something like Emacs which is so extensible, it is good to promote customizing it to one's own taste and preferences instead of installing a readymade customized distribution of it. This project is meant to be a quick starter kit to quickly do these customizations without having to spend too much time on it.
I can tell you why I love his essays: because of the way he argues. It is very transparent. I seldomly agree with his conclusion(s), but since he is so honest (he is not trying to deceive) in listing his arguments it is just a matter for me to find where he is wrong... or where I am wrong. I find this style so uncommon and stimulating that I hardly can stand the regular "opinion" held by most people. At least it is hard to be intrigued by arguments that doesn't even try to be precise.
When PG writes something I pay attention to the words, because I know they were chosen with some care. If I can force open a crack in his arguments I get wiser, and if I don't... the same thing happens.
I wish more people wrote like him, but preferably less about startups. They don't interest me as much as attempts at honest discourse.
What are his arguments? That's what I'm puzzled about.
Usually an argument is a series of premises followed by discussion of how those premises combine to lead to a conclusion. The conclusion can be deductive or inductive. I am not demanding a scientific or sociological proof - but there is not even an attempt at explaining why these premises hold. There are two types of people in this world - independent and non-independent thinkers. Why? What if there are 10 types of people in this world, all landing on a spectrum of independent and conventional thought mixed in different aspects of their lives? Why should I accept the premises? What good is reading a conclusion (how to cultivate "independent thought" in the second half of the essay) when I don't know if the premises are even valid?
Yes, but we also didn't close down our economy nearly as much as most other European countries. I'm just saying that there are too many variables to conclude anything at this point, especially given that we won't have a good guess on "final" death toll until widespread immunity (from vaccine or otherwise).
Why compare with Portugal? Why not compare with some other country with similar deathrate as Sweden but completely different choices? Maybe because you assertion about "very bad choices" then wouldn't hold any water.
It might still be true, but the actual support for it seems to just be cherry-picking.
Rereading my own post some hours later I don't like the tone of it, so I'll willingly appologise for that. This might warrant a FUD-claim (I had to look up FUD to start with).
The point I was trying to make, which I don't at all think is based on fear, uncertainly and/or doubt, is that any two countries will differ on so many variables that one can almost freely pick what correlation one thinks is the important one. To do something akin to that and the go "shame on you" feels wrong.
“We did not exploit any flaws or vulnerabilities. The cameras have a built-in maintenance backdoor, which allows anyone with super admin privileges to access a root shell on any camera of any customer at the click of a button.”
https://ipvm.com/reports/verkada-hack