[lukeschlather]

Another thing is that the native Windows app can support 50 simultaneous video streams, which makes it possible to see more people. I was in dance parties during the pandemic with enough people that it meant that I could see 50 realtime streams of people dancing. It wasn't the same as being in a room with 50 other people dancing in sync, but it was better than only seeing 25 people at once. My problems aren't so much nebulous as too numerous to explain, I could go on for a while.

Of course a lot of this is Zoom vs. Google Meet, I'm sure a lot of the things I like about Zoom work fine in the browser - but not as well as with the simultaneous video streams limitation.

You can cost out security, but a lot of the things that I love about Zoom's native app are truly priceless - it means I can see and hear more of people I care about. Another thing is supporting dual monitors with different screens, it makes it very easy to rearrange and see more than one person I want to see at a glance. You can do it with multiple browsers and so on, but it's more fiddly and you spend more time fussing with the screens, which means less time actually paying attention to the people.

[throwitaway1123]

We might simply have different value systems because I could never justify putting myself at risk of a data breach in order to badly simulate the experience of attending an actual social event and dancing with other human beings. That feels very dystopian to me.

We're getting really deep into the intricacies of Zoom and Google Meet here, and I feel like we're losing the larger plot. If you have a battle station set up for Zoom parties with multiple monitors with 50 simultaneous dancers that you need to keep an eye on, and you don't mind the security risks, then maybe you represent a specific edge case, but I think the vast majority of software users have different requirements that web browsers satisfy handily.

[lukeschlather]

You're at risk of a data breach the moment you connect your computer to the Internet. You need to do a complete threat model and explain how Zoom contributes to that risk, and weigh that against the benefits. If you have zero tolerance for a data breach you should delete the data so no one including you may access it. Zoom is reliably effective at transmitting data, you can use less reliable methods but Zoom deliberately often makes the choice that delivering data is preferable to not delivering data. I think this is a valid choice and in security we sometimes have to say "would I prefer to open myself up to attack, or would I prefer not to deliver this message at all?" Both are valid choices in different circumstances. Practically speaking I have conversations in public places all the time and I don't stress about the possibility that someone might be recording me with a parabolic microphone.

[throwitaway1123]

> You're at risk of a data breach the moment you connect your computer to the Internet [...] If you have zero tolerance for a data breach you should delete the data so no one including you may access it.

To reason by analogy, this is like me suggesting you wear a seatbelt while driving a car and you responding by saying: "well you're at risk the moment you step outside of your house, so if you really have no tolerance for injury you should simply not leave the house". You're saying instead of opening Zoom in the browser I should delete all personal data from my computer, and for what end? I'm doing this so that I can attend virtual dance parties efficiently? I don't understand how any rational cost benefit analysis could yield such a conclusion.

> "would I prefer to open myself up to attack, or would I prefer not to deliver this message at all?"

This is absolutely a false dichotomy. The choice isn't between sending data and not sending data, the choice is between sending data in the browser vs sending the same data within a desktop application.

> You need to do a complete threat model and explain how Zoom contributes to that risk,

The Zoom desktop clients have had RCE vulnerabilities where hackers were able to remotely execute arbitrary code on victims computers with zero user input required from the victims (they demonstrated this by remotely opening the calculator app). It's very obvious how Zoom contributes to that risk. "A zero-day vulnerability in Zoom which can be used to launch remote code execution (RCE) attacks has been disclosed by researchers. The researchers from Computest demonstrated a three-bug attack chain that caused an RCE on a target machine, and all without any form of user interaction [...] an animation of the attack in action demonstrates how an attacker was able to open the calculator program of a machine running Zoom following its exploit. As noted by Malwarebytes, the attack works on both Windows and Mac versions of Zoom [...] The browser version of the videoconferencing software is not impacted." [1]

> Practically speaking I have conversations in public places all the time and I don't stress about the possibility that someone might be recording me with a parabolic microphone.

Do you yell out your bank account number and routing number in public because you think the user experience of finding a private place to talk is too burdensome? Because that's metaphorically what you're arguing for.

[1] https://it.slashdot.org/story/21/04/09/209227/critical-zoom-...

[lukeschlather]

> The Zoom desktop clients have had RCE vulnerabilities where hackers were able to remotely execute arbitrary code on victims computers with zero user input required from the victims

There have been RCE vulnerabilities in browsers too. Do you have an example of a Zoom RCE vulnerability that wasn't fixed? The example you gave was one where Zoom was proactively publicizing their own work to recruit researchers to find vulnerabilities so they could be fixed before they caused actual issues - and Zoom fixed the issue, you're using Zoom's good behavior in security testing their app against them.

> Do you yell out your bank account number and routing number in public because you think the user experience of finding a private place to talk is too burdensome? Because that's metaphorically what you're arguing for.

No it's not, I wouldn't transmit my bank account number and routing number or similarly sensitive information over Zoom.

> This is absolutely a false dichotomy. The choice isn't between sending data and not sending data, the choice is between sending data in the browser vs sending the same data within a desktop application.

The choice is in fact between sending data and not sending data. I've given you one example (the limited number of simultaneous streams) where you're opting not to send data. You're just pretending that use case is invalid. There are other examples I could give, but they require more explanation and you seem determined to dismiss any examples I give.

[throwitaway1123]

> There have been RCE vulnerabilities in browsers too.

We're not comparing the security of web browsers as a whole to the security of the Zoom desktop client, we're comparing the security of the Zoom web app to the security of the Zoom desktop app. Can you find a single example of an exploit that allowed a hacker to execute arbitrary code on a user's computer after visiting zoom.com (even one that was eventually fixed)?

And also, almost all computer users are running web browsers (they come preinstalled on most consumer operating systems). So by downloading Zoom you're adding an additional threat vector on top of whatever threat your favorite browser already represents.

> The example you gave was one where Zoom was proactively publicizing their own work to recruit researchers to find vulnerabilities so they could be fixed before they caused actual issues - and Zoom fixed the issue, you're using Zoom's good behavior in security testing their app against them.

Every reputable organization has a bug bounty program (including browser vendors). You don't get a participation trophy for having a bug bounty program. You're missing the entire point which is that Zoom offered $200,000 to security researchers to find a vulnerability in their products and both of their desktop clients produced critical vulnerabilities yet the browser version did not. Which I would infer means that the browser version is more secure. Once again, do you have any examples of Zoom exploits in the browser?

Zoom can't keep producing vulnerabilities and getting a pass because they eventually fix them. This exploit's existence was publicized before Zoom fixed it. What if it was sold on the dark web and exploited in the interim?

> No it's not, I wouldn't transmit my bank account number and routing number or similarly sensitive information over Zoom.

People are absolutely transmitting critical information via Zoom. Courts literally use Zoom to remotely administer proceedings. I've heard of companies that ask employees to hold up sensitive documents in front of the camera to verify their identity. Hiding your head in the sand and saying "just don't send sensitive info via Zoom" is disingenuous at best. And even if you did buy in to the "just don't send sensitive data bro" argument it doesn't matter because an RCE exploit could potentially expose information that you never transmitted via Zoom that just happens to be sitting on your filesystem.

> The choice is in fact between sending data and not sending data. I've given you one example (the limited number of simultaneous streams) where you're opting not to send data.

The overwhelming majority of your examples have been cases where you've begrudgingly admitted that there's a way to accomplish your goal in the browser, albeit less efficiently. You started this conversation by admitting that you can share your screen in the browser, but doing so via the desktop app saved you a few minutes. I haven't found any published documentation from Zoom in regards to the streaming limit and I'm not willing to call up 49 other people to test this for an HN debate, but like I said, the vast majority of users can fulfill their needs in the browser. The average Zoom user is not hosting digital raves. I couldn't even imagine wanting to have 50 videos playing on my screen vying for attention. And again, there's no technical reason for why you couldn't implement 50 streaming videos in the browser. Maybe Zoom should spend that $200,000 improving their browser product.

[lukeschlather]

Efficiency is very important. The whole point of using a video app is to communicate things you can't communicate as quickly without video. It would be much safer just to use text, even voice would be significantly safer.

And again, this is about messages simply being undeliverable without the native app. If the web app takes an extra 5 seconds to join the meeting (I think it can actually be worse than this in a lot of cases - minutes lost) and I've just reached the meeting at 2pm and the meeting starts promptly at 2pm, and the native app causes me to miss 5 minutes of the intro... what is the cost there? I think in a lot of cases it can erase the benefits of video.

Courts are famously overloaded. If a court can get through 10 cases over a 4 hour session, each taking 24 minutes, and if each case loses 5 minutes of time due to using the browser app, then that's two whole cases worth of time they've lost. And it's not just about being able to handle a larger caseload: in a lot of cases inefficient communication results in incorrect communication and incorrect decisions.

Broadly though, you're taking it for granted that it's easier to make a browser app efficient than it is to make a native app secure. I'd actually wager there's a ceiling to how efficient you can make a browser app, and you can make the native app at least as secure as the browser app while also making it more efficient. Zoom has the money, they don't need to cheap out and do the browser app, and they shouldn't because efficient communication is a matter of life and death and their whole reason for being. Yes, it should be secure, but you can't just dismiss efficiency, being inefficient can cause security problems and it is often a matter of life and death.

> And again, there's no technical reason for why you couldn't implement 50 streaming videos in the browser.

Their docs specify that there are minimum system requirements for doing 50 rather than 25. This is an efficiency problem at the end of the day - the browser app is less efficient and can't handle as much info at a time with the same hardware.