YouTube Down
downdetector.com
downdetector.com
It's a strange feeling to be alive just after all of the significant technology companies were created and started to gain traction. YouTube has only existed for fifteen years, but it might continue to exist for centuries. The demand that has arisen in people to watch videos online is probably not reversible so long as maintaining a video streaming service is still physically possible. If not YouTube, something else would probably fill in such a void if it ever appears, barring the Apocalypse.
It gives me a lot of conflicting feelings. Twenty years ago, few people would have known they wanted the unique content that YouTube offered that wasn't available at the time, like livestreamers or swathes of content about every single plane crash or minor video game mechanic that can be talked about. But now, thousands have become hooked on them. I can't remember the details, only that I've watched them at some point. I wonder where the time would have went if we were not enterprising enough to invent big data and portable streams and engagement metrics. And for me personally, I can say that I "like" a lot of content, but I'm reasonably certain that's the kind of thing some portion of drug addicts would say about their habits. Not much of it actually helps me.
When a channel does well, the channel owner starts thinking, he can quit his day job and go fulltime as a youtuber. They borrow money and invest huge sums in building a studio. Then, they realize that they need to post not just regularly but often. You can feel that they are getting burned out. Sometimes, there will be a post titled "life update" where they will hint at it but will soon succumb to the fact they have no way out. They have to post a rate that is worse than torture.
Sometimes, Youtube gets upset or somebody takes offense and makes a complaint. The channel suffers a steep fall in traffic. The channel owner complains but gets no response.
Second, the good old days of "before" we were enterprising enough to invent big data etc. is retroactively constituted, and not apparent until its demise. Before YouTube it was television and radio, magazines and books, and even writing itself. On the technology of writing, Plato said "They will cease to exercise memory because they rely on that which is written, calling things to remembrance no longer from within themselves, but by means of external marks."
These constructions create the appearance of the industry's indestructibility -- certain consequences (the before, the help, or the livelihoods or subcultures you mention) are taken to axiomatic or self-evident, creating a logical construction that is bulletproof in its circular vacuousness. YouTube will die, just as websites before it have died, and technologies before it have died. What comes next might be better or worse.
That is to say, just enjoy your videos if they make you happy.
One, Veritasium. Content like this didn't (perhaps even couldn't) exist before YouTube. The science TV shows and even many books don't go into as much user-friendly depth as this does. You'd have to go to college and pay to get content even remotely close (and still, based on prior experience, not as good).
Two, any one of many How To DIY channels. They show you how to do things you're unlikely to find elsewhere. Where else will you find someone willing to tell you how to change your oil and filter on a [random year & model] car for free? How to wire up a light switch to code? How to clean your sink drain? Certainly not the paid professionals.
There are also channels like the Sampson Boat Company that take a project that would've just ended up turning an old boat into scrap after a while into one of my favorite carpentry/woodworking things to watch, better even than old episodes of the Woodwright's shop or This Old House.
The value of the digital equally does not imply a lack of value in the physical.
And, I will assert until I’m dead, the internet has done nothing but broaden my horizons - for the internet is full of quality information and entertainment, beyond what I could ever find in the physical world, if only because without a free or cheap publishing method the information and entertainment would not exist.
Some things are simply not possible with only a “physical economy,” because t physical economy is about things, not ideas.
> for free?
It's not really for free. People making these videos are getting paid.
> go to college
One thing I do appreciate is universities streaming their classes, on Youtube or other means. Of course, it's not that I will become an astrophysicist by just watching videos.
That said, there is also a lot of junk YouTube content, which feels similar to browsing Reddit, Facebook, or eating donuts for that matter... It's a guilty pleasure to watch some videos. Even the "cute" ones, like cat videos, or "funny" fail compilations, or impressive parkour videos. These can be entertaining on the surface, but I find they do the opposite of inspire me. They're temporarily exciting to watch, but overall do not any value to my life and in the big picture, I find them to be more of a distraction than anything. Some people have a balanced media diet and can consume it moderately. For me, it's too easy to spend several hours in any given week watching completely useless videos - especially with the ever growing mechanisms to try to keep your eyes glued to the screen (auto-play features, suggested playlists, etc).
I compare it to a type of drug use. Some people handle it well, others don't. Like most things, YMMV.
Edit: I also feel like with YouTube's insane growth over the past decade, there are way too many people trying to game the system. It leads to some very good content, but also a lot more junk content that just manages to add better keywords for searches, or a more attractive title/thumbnail to catch your eye (I'm sure many people are familiar...).
I was more meaning about being distracted. A lot of times it feels like browsing sites with a lot of updates is something I do when there is not immediately anything else to do. But that's an individual issue. If I am going to rectify that problem, then I need to ask what I would be doing instead, and do that, instead of falling back to whatever happens to be in reach. Sometimes I apply a rule of only using YouTube for X hours if I know exactly what I want to watch ahead of time, instead of using the recommendations to watch something I didn't necessarily know or care about but looks like an enticing rabbit hole to fall into from the presentation.
I do enjoy using YouTube as a platform, I just need better viewing habits.
Get a grip. Media companies come and go. They always seem unassailable at their apex.
But RKO is just straight up dead, which I don't think anyone would have called in 1943, its 15th year. 1946 would be its most profitable year. It was gone in 1957.
An example of a non-media company wanting to exist for 'centuries', I guess Evernote is dreaming to join the ranks of IBM, Nintendo, etc when they still want to be a 100 year old startup. [0] I think reality says other wise.
[0] https://www.fastcompany.com/3012870/evernotes-quest-to-becom...
There's also people deliberately misrepresenting knowledge for clicks & selling products. Micheal Crichton was right not to be concerned about the fall of journalism. He said (to effect) anybody can write whatever they want. The value is in the edited content. Which he meant by mainstream, validated press agencies that used to control most intepretation. The concept still applies, most of youtube will not be missed.
I feel like society is in the midst of restructuring itself to the reality that there will never be a single narrative again.
And I think that's ok because we're reverting back, societally, to a time before mass media. So we know it's possible for societies to function without a single overarching narrative.
(Not to go too far off, but the same thing seems to be coming as a result of deep fakes -- reverting to a time where 1st person accounts from a reputable person were the only reliable evidence. [Where reputation was something of an organic page rank algorithm.])
> AOL has only existed for fifteen years, but it might continue to exist for centuries.
Things change. No one can predict how, or why. But time marches on.
Although, when AOL was around you could only practically use the Internet from a large desktop machine, but now you can access a massive amount of information from almost anywhere using a smartphone. And we're being introduced to computers at a younger age, and are spending more time with them. But that might not be indicative of what happens in the future.
Then there is the obvious “the algorithm can dictate what you watch” that tik tok was so critiqued for.
But what would that form of media be, and what would be the justification for creating an entirely new media platform to succeed YouTube? I find it difficult to imagine what it would look like at this point in time.
People made a big deal about archiving Usenet when everyone started moving away from it. If YouTube has to be archived I don't see how it can be completely accomplished without people who have used youtube-dl in the past gathering up whatever they happened to save before it went down. There's hundreds of millions of hours of video already on YouTube, and unlike Usenet archives only a few large entities have the money and infrastructure to warehouse the entirety of that data.
Saying that YouTube will pass on because media companies always pass on almost sounds as if YouTube is a ticking a time bomb, and after enough time passes we'll inevitably lose a lot of things that brought us value en masse, and the scale of loss will be far greater than in the past because of the breadth of content and the sheer ubiquity of YouTube.
This is an excellent point. With the multitudes of content competing for our attention, it's important to be very selective and deliberate with usage. Otherwise, it can very easily turn into an addiction - you'll be mindlessly enjoying the dopamine hits of watching "interesting" videos whilst life passes by.
Is YouTube profitable yet?
I really doubt YouTube will last for centuries, or even twenty or thirty more years. There was a point where it seemed like Geocities was ubiquitous and would be around forever, and now it's gone. Internet services don't have a great track record for longevity. When YouTube finally goes kaput, much of its content will likely be lost, since I don't see Google going through the effort of offering archivists a dump.
Though, I'm kinda skeptical that even much digital data will last for centuries, unless formats and storage media ossify at some point on some long-term stable standard. The trend towards cloud storage makes things even more precarious.
Google doesn't disclose that. Only that Youtube have revenues of $15 billion last year.
Now if HN were down...
Neither do TV or movies for the most part; there are many examples of high art in those formats but the most popular examples thereof are hardly at all intellectually gratifying, educational, or otherwise "helpful", as you put it (a good descriptor). It's entertainment - it's there for fun, and in many cases, it's there for the sake of art. (That's not to say that e.g. popcorn flicks don't have any artistic qualities, just that they're aiming for the former, not the latter.) The issue is more in the ways that YouTube tries to keep you watching, like the other massive social media sites.
No one else, ever, has had a platform where an independent video creator could start with a handful of viewers and build up to millions, and even potentially make a career out of it. No, it's not common, there's a ton of work involved with no immediate payout, and it requires at least a bit of luck with the dreaded "algorithm", but it's probably more merit-based than what we had before, and that's a wonderful thing.
(I'm not entirely certain about that last bit, hence "probably". How merit-based was/is TV and film, precisely? I just find it hard to believe that they're better than YouTube, given the quantity of quality creators on it who could never find a slot, much less an audience, on any other platform or format that has ever been.)
With regard to the ubiquity: given the lack of competitors, apparently the capital requirements are so high and the lock-in effects so strong that there aren't any companies willing to give it a go. The closest thing we have, Twitch, filled a niche that YouTube wasn't technically capable of at the time, grew large enough before YouTube decided to try to enter that market so Twitch was able to acquire those same lock-in effects, and was bought by Amazon who provides the capital to ensure a smooth experience (Twitch prior to the acquisition was far more prone to performance issues).
Out of habit, I do go to youtube if I need to search a video of something.
Do remember channel surfing on your dumb, terrestrial, flow TV? Hours spend clicking "next channel" on the remote, cycling the same 5-10 channels, passing by the same program, again and again.
That's where the time would have gone.
(I have no idea whether YT revenue is meaningful for smaller players, to be clear; I suspect not, but I also suspect it's worse elsewhere.)
There's a 2018 copyright date in the footer, so I guess that's as alive as it needs to be.
I noticed the outage because I wanted to watch a stream, so I might go to Twitch now. Often I use Youtube for music, where Spotify is a solid alternative with many of the same creators. If I want to see that video from reddit of people throwing burning flares at a balcony I can probably find it on LiveLeak. Many of the educational channels I've subscribed are also on Nebula, but I could also watch a documentary on Netflix or Disney+ instead.
When your main differentiator is "free speech" and your competitors don't actually restrict the speech they carry that much, you're pretty much only going to get the dross that was cast off the other platforms and very little else.
But Youtube doesn't only ban conspiracists and hate speech. They also will restrict content that violates their community standards such as explicit instructions on slaughtering chickens on a farm, or how to field-dress a deer when hunting.
Even the Rob the Ranger channel has had videos taken down for inappropriately explicit content, when showing lions mating for example.
Youtube throws a wide net, and thus rather unfairly censors innocent content. Bitchute, peertube, etc. provide alternative ways to get that content online while you're waiting patiently for your appeal to go through.
And yeah, there's some content that the monitors at Youtube consider objectionable and will never be allowed, and maybe should never be allowed -- snuff films, child porn, terrorism how-to training guides, stuff like that. But I think Bitchute also takes down such content.
Edit 45 min later: Everything appears to be working again, including YouTube TV.
(Does anyone still remember when you could easily download videos from YT by simply replacing "watch" with "get_video" in the URL? I miss those days... when corporate greed hadn't gotten to where it is today.)
But it makes it hard to download, unless you stitch all that on the receiving end, something that youtube-dl does.
Greed and etc. were already piled on top of that with DRM, obfuscation and the like.
The supposed excuse of that overreach in giving the Librarian of Congress the ability to define exceptions is totally lame and unacceptable.
This was a change that I am assuming was added before any complaint could be mounted about this use case for fear of striking more of the DMCA down than just that provision that was modified.
Youtube likely has lots of moving parts to host all the different systems. Video content, search, browse, player, etc... could all be different bits. And it's also fun to know that video content itself may not be hosted directly by Google, as you could get it from the GGC (Google Global Cache)[0].
When you have a system as big as Youtube, you need to think about all the different ways you can slice different parts of the service. If any of those parts could be more optimized by having the data stored/sorted in some different way, it likely is.
[0] https://support.google.com/interconnect/answer/9058809?hl=en
Search ''Meme text' knowyourmeme' on you search engine and you'll get an answer.
"Youtube is asleep, post Vimeo" and a link to the old, old viral numa numa video hosted on Vimeo.
[1] https://imgur.com/a/jDYji8s [2] https://www.reddit.com/r/videos/comments/jsky61/youtube_is_a...
But the technology isn't the problem. The problem is nobody is going to use your platform. We need to decouple discovery from hosting. We need a slick, simple aggregator that lets people submit video URLs from multiple platforms, then handles recommendations. If your video has an equal chance of blowing up regardless of where it's hosted, it opens the door for competition with YouTube.
A useful app that I can think of is one that supports subscribing to various independent video publishing websites... but aggregates them in one app.
I think those 2 are actually leading the pack because they consider themselves technology companies.
i'd wager some database storing info about the streams crashed :^D (given that it also crashes on /, it's more likely something else)
What’s going on?
edit: I assumed you meant more in general over the past weeks or months, not right now - if that's the case.
But those are indeed some interesting patterns, that so many websites are getting the same spikes.
Doesn’t Downdetector count only users who click on “I have a problem with [service]”? It would seem dumb if all charts moved together any time there is a major outage of a single service.
I just ran into that issue trying to download a few different apps an hour ago but figured it was my phone. Something is definitely up, too much of a coincidence for these issues to be happening at the same time
Down detector scraps sites (twitter and reddit) for 'website.com down' and marks a spike
> YouTube ads generated $15.15 billion in revenue in fiscal year 2019.
So plugged into wolfram alpha:
https://www.wolframalpha.com/input/? i=15.15+billion+dollars+per+number+of+seconds+in+a+year
They lose about $480 per second or $1.7MM per hour. Yikes.
But I think in this context MM is a million.
So:
(number of new users) x (probability of those users not coming back) x (lifetime value of a youtube user) = I highly doubt this is equal to 0
videos however come from redirector.googlevideo.com/videoplayback
Perhaps related, I uploaded a video earlier today that took nearly 4 hours to process, though it was only ten minutes long.
[0]: https://old.reddit.com/r/youtube/comments/jsjwzq/youtube_dow...
"We're so sorry for the interruption. This is fixed across all devices & YouTube services," YouTube said in a tweet https://bit.ly/36r4sjz, without explaining what had caused the outage.
Google did not respond to a Reuters request for comment on the outage.
https://www.reuters.com/article/us-youtube-outages/youtube-b...
YouTube's tweet: https://nitter.net/TeamYouTube/status/1326709701526974464
Would love your opinions/wish list/must haves.
Distribution has always been historically expensive for any content. While digital distribution is cheaper than ever, B/W costs for a product like YT is enormous .
To make unit economics to any sense, you will need to spin your own CDN and buy b/w from other CDNs when that it is cheaper. For your own CDN, you are looking at few hundred PoPs and thousands of peering agreements with ISPs in each geography maybe lease cables too . This is a substantial investment.
Even if do all this you still have to compete with Google on android where they have a distinct advantage, YT is pre-installed in billions of devices and user account already activated. You will have to pay OEMs a lot of money to get the same kind of user base pre-installed and still do your own user activation.
There is a reason why nobody is trying seriously. the next largest competitors DailyMotion and Youku are 1% of YT size.
Many people have pointed out issues with recommendation algorithms, both on YouTube and elsewhere—particularly the “falling down the rabbit hole” phenomenon of getting deeper and deeper into conspiracy theories. I generally stay away from politics on YouTube, though, and most of my recommendations are actually quite helpful: good videos on topics that I’ve been viewing recently or have searched for.
So, for me at least, alternatives to YouTube would have to have those two features at a minimum: seamless uploading with few restrictions, and good recommendations. Whether that would be feasible commercially I have no idea.
Oh, one more thing: Hire the HN guys to design the moderation system for the new service's comments. YouTube comments are, in general, horrible.
That would be sweetly ironic, considering Google's original success came from better search.
Will be looking forward to their report.
Usually this type of site-wide outage is always related to networking.
BGP, not sure. Google practices SDN. So their software error can rekt a large chunk of infrastructure...
I'd want to know what caused this. I'm guessing they're running on kubernetes, but it too early to tell what actually must have failed..
clearly there are points of failure and not sufficient backups to that point of failure..
The play store appears to not be serving any downloads. Sounds like some serious CDN issues.
I think it's load balancer related for whatever is serving their videos
EDIT: Once you have a connection you appear to keep it the entire way through (at least it did for me with youtubedl). Something related to connection queueing...
I actually get back a little webpage it says:
502. That’s an error.
The server encountered a temporary error and could not complete your request.
Please try again in 30 seconds. That’s all we know.How old is this? I think I've always seen this error message from Google. It's at least 15 years old.
There wasn’t anything unusual about my CPU behavior when this happened, I can only assume it was GPU or some other part of the system (unfortunately I was too busy to do more than basic troubleshooting at the time.)
Based on a recent 15B/year YouTube revenue figure I found, that would mean ~$500/second.
Unless BitChute is hosted in Google Cloud and this is just a Google Cloud problem, I have a feeling something major is being censored.
[0] https://www.businessinsider.in/tech/apps/news/instagram-down...
Perhaps related, I uploaded a video earlier today that took nearly 4 hours to process, though it was only ten minutes long.
Just a guess.
Now to be fair, When in a meeting, I’ve al had to track down the kids and all devises that could possibly access YouTube.
After a really bad hour of Zoom; I leaned that you can watch YouTube on the switch. Clever kids.
I locked away a 10 year kindle just out of principe for that one.
YT have been the only player in the game for far too long.
Edit: looks like it might be back up now.
EDIT: just kidding - playback is intermittently hanging now.
EDIT: and its all back up to working
Edit: Err, probably more serious
I wonder if Google is have their moment :)
seems like these web services always have to go own eventually.
Videos viewed via website since it went down: 0
Videos downloaded via youtube-dl since it went down: 2
(Granted, there were plenty of retries by youtube-dl.)
Probably...
Seattle.
EDITs: More specific in location.
If it happens, though, my paranoia is going to climb dramatically...
And there are also the videos where the "original copy" was on YouTube itself: unless you have a setup to also record locally (or mirror to another site like twitch), streaming videos are only recorded on YouTube.
Now I know why I used to download all my favorite videos to USB/SSD!
But, the video you are trying to watch won't load, and will throw a 500 error. But this isn't considered downtime, for them.
Yeah I've had nothing but grief with those outage detectors. They seem to be really optimistic and not take into account the web of CDN's that contribute to the 'uptime' of a website