Comment by btown | Hacker News Reader

Comment by btown | Hacker News Reader

Parent Full thread

A welcome move for individuals who may have embarrassing content as Unlisted links. Future politicians will thank you. But... this will hit B2B product training and product marketing libraries hard. Many companies I've seen have help pages with embedded or linked videos for features not updated in years, and many of those embeds are Unlisted videos so that they're only seen in the context of their help article, not promoted randomly by the YouTube algorithm. Some may have legacy content on legacy "X Corp Training" YouTube channels where nobody knows how to opt out of this policy shift. And especially post-COVID, they may no longer have the same technology and training teams, if they have any at all. They may not even have the YouTube login.

I could see a policy where YouTube made Unlisted videos Private that only had referrers from social media; this would be a welcome compromise to ensure non-guessability of URLs. But I can also see how this could become complicated and political. And companies using YouTube in this way aren't really contributing to YouTube's revenue materially, so there's not much incentive relative to the reputational risk of people guessing Unlisted links.

I shudder to think that healthcare professionals or heavy-machinery operators might be relying on these links to be trained in systems they use, will start to see broken links, will never report them back to the right people at their system providers, will just not get the full training, will make mistakes, and might cause harm as a result.

Security is not the only component of safety, and impacts need to be evaluated holistically.

So your complaint is that a company might have put a forklift training video on youtube, as an unlisted video, have then forgotten the password to change it, then will hire someone and tell them to watch it, that person won't be able to and will then not tell anyone else that fact, and after all that they might then hurt themselves with a forklift?

At what point is a hypothetical harmful scenario too absurd to care about?

This is the best forklift training video I have seen:

https://en.m.wikipedia.org/wiki/Forklift_Driver_Klaus_–_The_...

YouTube link (not unlisted):

https://www.youtube.com/watch?v=IAHzP4umE4M

Who says germans don't have a sense of humor (OK my austrian friends do, but you know those austrians...)

My old university prof used to say: "How many Germans do you need to change a lightbulb? One, because we are efficient and don't have time for humor." (German university + German prof)

Yeah, the Austrians are saying that, but here in Germany we say the same about them.

Austrians are the world's sneakiest bastards:

They managed to convince people that Beethoven was Austrian, and Hitler German.

The factory where anything that could go wrong does go wrong.

"Although the film is not officially part of the German training and education system for forklift trucks, it is frequently shown by instructors to lighten the mood."

I was born in the wrong country.

I know I would be an impeccably safe forklift operator after watching Klaus' training video. :)

This is amazing. Both a good forklift training video and a hilarious award winning short film!

Thank you.

Thank you for sharing a treasure I somehow missed until today. Would enjoy this style applied to other safety videos!

Oh wow - it just gets more and more gory! Hilarious.

Thank you for this.

The scenario you described doesn't sound absurd to me at all.

Even at high-functioning companies, people don't report or fix glaring problems in documentation and training material. There are quite a lot of not-so-high-functioning companies in the world.

I'm confused by your confusion. I am absolutely positive I have made videos at old workplaces that will be impacted by the change. Who knows if they'll figure it out!

I think the absurdity is that the videos you made are necessary to avoid injury?

And also that when the links stop working you just skip that part of the critical training rather than ensure the safety information is conveyed some other way.

I mean, what's the flip side complaint? That a person posted a video that they didn't want to share, but as Unlisted, then forgotten the password to change it, then the video will be discovered, and it's compromising information but was uploaded to YouTube, and it comes back to harm that person?

They're both pretty unlikely but at YouTube scale both are real.

It doesn't seem that unlikely. Someone makes a video of themselves with no clothes on (or whatever) and not fully understanding the privacy options they make it so it can only be shared to people they choose.

15 years later it's much easier to trawl through millions of videos by incrementing and check them automatically for nudity than anyone thought it would be back in the day.

Obligatory: https://www.youtube.com/watch?v=gocwRvLhDf8

There’s too many IDs to iterate through. Searching for $defatltvideoname for some camera is far more effective for that. Scarily so.

> At what point is a hypothetical harmful scenario too absurd to care about?

They’re YouTube. At their scale, there’s no such thing as an “edge case.”

Nah, they still have weird edge cases. That's what keeps eg SREs busy. (I used to work as an SRE on Google Photos.)

But at YouTube scale, the cases need to get weirder and weirder to fall into 'not-yet-handled' territory.

Sometimes all the holes in the cheese still align.

That sounds extremely feasible to me. One of the companies i used to work for will definitely have a bunch of their training videos lost this way, and it's totally reasonable to believe that a new hire for the warehouse would not bother telling anyone that the video wouldn't play. I don't know that any real injury will result from this, but i wouldn't call the scenario absurd

Its about responsibility, not absurdity. It is Youtube's responsebility to protect customer privacy, wheras a company's management is meant to take responsebility for staff training and use appropriate tools/hosting for that.

Most people when confronted with a problem, just ignore it.

This can be hard to imagine for a problem solver personality. How can people be so helpless?

But it’s basically because people have learned that complaining usually achieves nothing so they don’t even bother.

when multiplying it by youtube size isn't still a big number?

Well then you've got a decade of free hosting and the only negative consequence of it is it stops working.

The reductive logic here is something else

> A welcome move for individuals who may have embarrassing content as Unlisted links.

All those "early access" on Patreon.

Maybe critical training materials should not be published to YouTube?

There is no better way to provide video content with a text document than to link to youtube. Sure they should have a private backup but until youtube stops working well like it has for a decade, it's still more reliable than any self hosted setup a company would build.

There is no better way to provide video content with a text document than to link to youtube

That may be the best "free" way, but my company puts their training videos up on some training website. The site tracks which videos were watched and sends nag emails when required trainings are late.

So if you're putting up content that your employees are obligated to watch, there are better alternatives than Youtube.

But then the video viewer doesn’t have the necessary features. Especially in 2010 when those videos were put up, everyone would have used the superior Youtube or Vimeo player. But they would have chosen Youtube because you had no guarantee Vimeo would be correctly paid for years.

There absolutely is, place it on company google drive! You will even have real access control!

Last I checked, google drive videos and youtube were in a weird merged state.

Dropbox,OneDrive, etc. If you want

Shouldn't it be a self hosted server?

The Google drive could be deprecated at any moment

The chance of your self hosted server going down, being hacked, catching fire, etc is infinitely more likely than google drive being shut down so fast you can't switch to something else.

At this point the advice is moving towards "Don't even have a video if you can't be bothered designing a CPU from scratch to host it on".

Keep a local backup by all means but it's far more likely the local copy will get lost than the google drive copy.

I'm a little skeptical of that, especially in the scenario that it's a "don't want to look at it much" situation. I've had VM's running for many, many years in an autoupgrade scenario, and it tends to just work. And the security risks are often overstated as long as you tried to minimize attack surface; so; for instance I for a while tried to track all related security issues - and at the very least 95% are irrelevant, and that last 5% is more CYA uncertainty than actual belief its relevant; I didn't find anything positively certainly relevant.

Obviously, if your website is running a complex stack this is fairly infeasible (good luck keeping a complex web-framework secure without vigilance!) But if it's just a static site... and static sites are the fair comparison to google drive.

The issue with google drive and similar products is that they're actively - very actively - maintained. Stuff doesn't just break because it's taken offline, but because there are changes in terms of service, there are automatic deactivations for inactivity, there are plain old incompatible upgrades, etc, etc, etc.

I mean - I totally support people starting with the minimal effort (which is surely some freemium service like google drive), and leaning to live with the limitations - but the risks of a server (whether physical or otherwise) if you're similarly conservative and willing to live within constraints not unlike hosting on drive seem overblown - people talk as if servers burn down all the time, and hackers crack everything, but nothing could be further from the truth - those are exceedingly rare circumstances. And for many things, those risk are just OK. Don't keep your only copy of the data on the box, and don't keep private stuff there, and if lighting really strikes - it's not so bad. And both strategies will need a little TLC once and a while, that's just the way it is.

I have seen plenty of these servers left unattended for years and they all build up weird issues like time drifting out of sync, log files filling the disk, hackers somehow getting in and setting up crypto miners.

None of this is stuff the average company needs to deal with when youtube works just fine. I'm sure you can list out all the ways to avoid those listed problems but that just shows how much work and knowledge is needed to pull this off while anyone can use youtube.

> And the security risks are often overstated as long as you tried to minimize attack surface; [...]

You know that this requires some uncommon computer proficiency?

Sure, and I certainly would recommend people start with an off-the-shelf solution like google-drive. But that recommendation is based on precisely that - you don't want to need that expertise, nor risk screwing it up. But it's not based on the fact that it's actually pragmatically more reliable (or less ongoing effort) to host on something like google drive.

Yes but self hosting is just so hard it’s cost effective for a lot of corporate users to freeride on social networking in general

Self hosting is as easy as having the videos on a single location on the shared drive. Most companies that have employees use computers use internal shared drives like this to share files and VPNs for their employees at home to access them outside of the network, and have their own IT departments to handle all of this.

YouTube is probably a more reliably backup in the vast majority of cases than whatever internal backup the company would come up with.

How can anyone rely on a third party business to host content for free, forever?

Just because it’s worked for years doesn’t make it so forever.

The trust and reliance on Google for free video hosting seems both foolish and dangerous.

Nothing is forever, including your own copies. Put it on youtube and keep a local backup. Most likely the local backup will get lost before the youtube one does. It's most likely the youtube copy will last longer than your company does.

Unless you are running a sole proprietorship, any business is composed entirely of 'third parties'.

See https://en.wikipedia.org/wiki/The_Nature_of_the_Firm for a classic take on the subject.

That may have been true, but these days can't you just put the video on S3 or any storage and play it using the <video> tag?

If you're a nerd into cloud providers, web development, etc, then mabye that sounds reasonable. If you're some person in charge of training that has none of that techy background, putting a video up on YouTube sounds like the perfect idea.

Yea, that is entirely doable. Some transcoding might need to be done first though. The worst part is that you’ll have to get your IT department involved.

> The worst part is that you’ll have to get your IT department involved.

That's also one reason spreadsheets are still so popular:

As an average corporate drone, you can either hack together something quickly over the weekend in Excel that will mostly work. Or you can wait 3+ months for your IT department to fail to deliver.

(And that's the best case scenario.)

You probably can, but the people who are tasked with delivering training videos likely do not have the technical ability to do so.

Sound like they do not have the skills or tools needed to do their job.

They do, they made a training video and put it on youtube. Now after years of benefiting from that for free they have to press a button to make it keep working like it always has. Seems like a pretty good and reliable setup.

They do. S3 and <video> tags are not usually part of that skill set.

In a world _no one_ has said skills and tools anyone becomes auto qualified in relative sense

This is the problem with these services.. they are free and every one was using them without thinking about what they 'should' do, because they didn't knowing any better, didn't understand that the service could change at any moment.. and we lost a generation of important records because people stopped posting on their websites and blogs, and started posting on instagram, twitter and youtube

Or they were, but they had to do more with less. Given how companies love to "cut costs," the dedicated training people who had the time and expertise to think about how the task "should" have been done may have been canned. However, the need still exists, so the task fell on some overworked person with other responsibilities to juggle, who barely has enough time to get a quick an dirty job done with Youtube.

I believe management consultants call that "efficiency."

Its local governments that I worry about more than businesses.. One my favorite nature reserves had almost 20 years of monthly photographic records on a blog that was killed by a switch to instagram.

"should not be published" is very different from "are not published"

agreed. YouTube as a platform has shifted away from hosting to "publishing" and edge-cases like this are unfairly tossed to the side.

coperate video training platforms are mega dumpster fires caked in the legacy user experience crust you would expect.youtube ends up being the best, most user friendly platform for companies to upload videos to. not to mention its free so "why spend money on this platform or spin up resources for self-hosted infra when this is free?"

There was a time, when YouTube was seen as free and open as Wikipedia.

There was a very long time when YouTube bled absolutely incredible amounts of money.

Maybe! But they were.

IMO: it's always been pretty dumb to use youtube for important things, doubly so important things that avoid the discoverability aspect.

Reply on news.ycombinator.com