We Call on FOSS Contributors to “Exit Zoom” | Hacker News Reader

We Call on FOSS Contributors to “Exit Zoom”

sfconservancy.org

95 pointsthedeepself48 comments

As a Deaf person using sign language, Zoom is essential for virtual meetings as its the only platform I've used that consistently delivers acceptable video performance. Teams suffers from random issues that just crops up out of nowhere eg videos going blank, sudden reductions in bittate, periods of excessive dropped frames, etc. Google's platform (whatever its called these days) struggles to cope with multiple videos streams, descending into pixelated messes due to adapting down to incredibly low bitrates and regular dropped frames. Zoom is the only one I've come across that manages up to 25 simultaneous video streams without visible dropped frames and at reasonable resolutions and bitrates.

Am keen to know what realistic alternatives there are for virtual meetings for sign language users, delivering quality on par with or at least 'close enough' to Zoom. If theres a OSS solution that can do that out of the box it'd do very well indeed.

I do software contract work for many blind and low vision people, and they also have had great things to say about Zoom's accessibility.

Can you try and comment on

meet.jit.si ?

Jitsi is quite a nightmare to work with, and presents a whole range of challenges and difficulties to use and support. Users often encounter difficulties with connecting, experience unexpected participant drops, struggle with subpar video quality, and face issues with loss of audio and visual streams. Unfortunately, there's also a complete lack of any adequate support tools, making it near impossible to address these concerns effectively.

The codebase of Jitsi is quite a tangled tightly coupled mess, exhibiting poor coding practices in Java. As a result, its performance is subpar, and it struggles to scale efficiently, especially in larger conferences. Even achieving consistent performance in smaller gatherings can prove to be a challenge. With respect to deployment, thats also another complicated mess. Browser version upgrades also often completely break the platform, requiring waiting for fixes then having to update and redeploy your entire system end to end because, as mentioned above, its very highly tightly coupled. You simply cannot compare this to native apps like zoom or teams.

While Jitsi might suffice for casual interactions among a small circle of friends, it pales in comparison to more reliable platforms like Zoom or even Discord. For those seeking a dependable alternative to platforms like Zoom, don't even bother with Jitsi as it falls far far short and will be a massive time and support money sink for your corporation.

Same with Slack, stuttering video in their 'huddles', broken audio.

Zoom works.

Google Meet performance is pretty much up there for me these days. It's very rare that quality degrades for any of the use cases you described. Maybe it's gotten better, who knows.

It has gotten better — still noticeably worse than zoom

It’s gotten much worse for me. Usability and quality peaked for me when it was just a wrapper around webrtc, and allowed anonymous users.

Now it requires an app and google credentials on iOS, and the video quality seems worse to me.

Google Meets is far behind- even image quality on 1-2 people call is quite low.

We must be using different versions then. We're using it for work, could there be a difference between free and paid versions?

Isn't sign language less efficient than typing? Are there conditions where you are on a video call with a keyboard in front of you, where you prefer to use sign language over typing? Or is the use case when you're away from a keyboard? Would it help to have text-to-speech built into the video software?

My understanding is that deaf people far prefer signing. For deaf people it’s easier to both sign and to read signs than writing.

Language emerged from gesture and it’s these people’s native form of communication.

What's your point? The same can be said with regards to oral communication.

People who communicate in sign language do so in a very expressive way, a way that pure text cannot capture. You should go out of your way to experience this for yourself.

My point is that I'm ignorant on the subject but interested and so am asking questions to learn about it.

Do you use audio/video chat? Why don't you type instead?

I personally prefer text to voice in most situations, and work intensively with several colleagues who feel the same at least for work. But I'm an introvert and recluse and recognize that this isn't normal.

And you won't get any enlightening information via text chat like this. Go experience sign language.

The person asked a question with curiosity and you chose not once but twice to respond rudely.

It is ableism wearing the mask of curiosity.

Sign language is not English.

I haven't had any issue using WebEx for large numbers of people. Jitsi is another option that has performed well whenever I've used it. (But I haven't tried Jitsi with huge groups.)

The OSS solutions include: Jitsi, Bigbluebutton (BBB) and Galène. I toyed with Jitsi when the pandemic hit but was battered into using Teams instead. It was very quick in my estimation.

Teams and Zoom suffer from (hyper)scaling issues. They have way more resources than you and I but when lots of people decide to use it simultaneously then it will suffer - death by a thousand or millions of cuts.

If you can self host on a reasonable platform then you will almost certainly get a far better experience than any big jobbie. You will need suitable internets for your server too. You will want 10-30ms latency and roughly 5Mbs-1 per attendee for 1080p. So for 25 users, you will need something like 150Mbs-1.

For you, I would get your mates together and self-host. Surely you have a nerd who could run the thing for you? If not, then farm it out but do be careful.

Obviously, where you actually are is rather important. I'm in the UK and will have rather different advice for a substantially different environment.

You are proposing a lot of work for someone who already has a solution.

"Do a part time job as an IT admin" is not a great answer. "Here's a hosting company that charges a reasonable price for a reasonable TOS" would be much more usable

I hope this isn't rude, but as a hearing person who knows very little about deaf culture, could you please help me understand why signing is preferable to text chatting?

And in a video chat with talking speakers, would you still prefer a sign interpreter, or are automatic captions good enough in that case?

I don't doubt your experiences, I just want to better understand them. As a hearing person, I often find it hard to understanding what people are saying over video chat (of any platform) and much preferable text chats or auto transcripts. But I'm sure that's not the case for everyone...

As someone who can hear fine and knows basically no sign language, that feels a bit like asking anyone why they prefer video chat to text chat. Text chat is almost always just slower. I can have a day going can and forth with someone over chat, or I can call them up and get the problem sorted in fifteen minutes.

Obviously there's still cases where the discussion is small and specific enough to do over chat, or other cases where slower, asynchronous communication works better because I went to do other things at the same time, there's no one-size-fits-all communication method, but talking (be that via English, ASL, or whatever other language) is very useful to have in one's toolbox!

A whole lot of people who were born deaf can't read English well.

But even if all participants in a conversation can read and write English, when your native language is some kind of sign language, writing in English is akin to speaking a second language would be for an English-speaker. Possible, but not nearly as convenient as using your native language.

Has anyone made the joke to you that "you're on mute" because the camera is off?

I do not wish to sound insensitive but is better accessibility the most important consideration in FOSS vs proprietary? The article covers the dangers of certain types of software and why we should not rely on them. It seems to me that particular features in the software itself is beside the point.

> A recent analysis showed that it could take up to 30 hours just to read the entirety of Zoom's terms and conditions

I wonder whether there's a single human being on Earth who has read the entire T&C word for word. I assume multiple authors drafted and edited specific chunks, so it may even be the case that even the lawyers who wrote it haven't read the full thing, yet users are expected to have done it in some vague but legally binding way.

> it may even be the case that even the lawyers who wrote it haven't read the full thing

idk. I thought lawyers loved billable hours. Maybe some poor intern at some law firm had to spend a few months reading the full thing over and over again until it all made sense to him :p

If it takes 30 hours to read, it will probably change before you're done reading it.

And how legally binding those T&C really are anyway?

Probably quite binding, unfortunately. https://en.wikipedia.org/wiki/Clickwrap#United_States

They didn't even quote that analysis correctly. It shows that it would take 30 minutes for the average person... big difference there.

I actually had to read this whole thing twice because the first time I thought they genuinely wrote all that without even suggesting a single FOSS alternative to zoom. They don't actually suggest one so much as effervescently mention one (BigBlueButton) near the end without describing it at all or even indicating that, yes, this is the what they want people to drop Zoom for. No wonder there's no author listed, this is one of the worst attempts at effective communication and persuasion I've ever seen. You can't sit there and say Zoom sucks without telling us why the alternative is better, especially when 95% of peoples' experience in the matter is that Zoom is an order of magnitude better than everything else they've tried.

We in the FOSS community need to stop being worse than our stereotypes.

Coincidentally, difficult to understand and largely ineffective is exactly how I'd describe the FOSS virtual meeting software I've used in the past.

That is possibly the least coherent press release/call to action I've seen in a while. If there is a convincing argument to be had there, it's buried in twelve paragraphs of flowery language.

"The proprietary and for-profit nature of Zoom also has made it subject to multiple cases of algorithmic bias. The once esoteric seeming issues are now a stark reality."

As a huge fan of FOSS, I gotta say this is ridiculous to include here.

Why stop at Zoom? Exit Discord too!

Remember when FOSS community discussions occurred in archived mailing lists crawled by search engines? How quaint!

Now your average project invites everyone to their "Discord server", the discussions equivalent of a black hole.

Strongly agree. Doesn't have to be mailing lists, but any persistent forum is better than discord

I will once there is a federated alternative with reasonable feature parity that I can operate an instance of on AWS.

I’m not willing to play security/regression whack-a-mole, so the server needs to be under 10k lines of code, have minimal dependencies and be written in a statically typed language.

Discord replaced Mumble, which arguably is mostly the same product, just less flashy (and no browser rendering)

I use WebEx, but I can use my video conferencing equipment to call into Google Meet, Microsoft Teams, and Amazon Chime because they are all interoperable. This is not the case with Zoom.

Standards exist for video conferencing, but Zoom refuses to work with them simply because they think their market position will allow it.

I thought Jitsi was the FOSS alternative. Surprised there's a new FOSS contender.

BBB has been around, and usable, since at least 2013.

Sure, when there's a consistent experience between browsers with decent echo-cancellation amongst other audio processing I'll be happy to move.

I don't want to get stuck on lowest common denominator WebRTC, that pales in comparison with a native application like Zoom.

Reply on news.ycombinator.com