Author Topic: eevBLAB #41 - VidMe Is Shutting Down  (Read 1902 times)

0 Members and 1 Guest are viewing this topic.

Offline EEVblog

  • Administrator
  • *****
  • Posts: 26841
  • Country: au
    • EEVblog
eevBLAB #41 - VidMe Is Shutting Down
« on: December 02, 2017, 11:36:50 pm »
Youtube alternative Vid.Me has failed and is shutting down December 15th.
Uploads disabled effective immediately.
Dave's thoughts on this and the Youtube adpocolypse...


 

Offline NiHaoMike

  • Super Contributor
  • ***
  • Posts: 4456
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: eevBLAB #41 - VidMe Is Shutting Down
« Reply #1 on: December 03, 2017, 01:17:15 pm »
Maybe they should have gone with P2P or hybrid P2P for content distribution. It has proven itself to be scalable with little or no need for a centralized resource.
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline Razor512

  • Regular Contributor
  • *
  • Posts: 132
  • Country: 00
Re: eevBLAB #41 - VidMe Is Shutting Down
« Reply #2 on: December 03, 2017, 01:27:02 pm »
Since YouTube is not profitable, a competitor that can actually be successful, will need to work towards a lower cost way to distribute the content.

A torrent based YouTube competitor can work well if made user friendly, especially since the largest cost aspect of a service like YouTube is the cost of distribution, and not the storage.
While storage can be expensive, it is not as expensive as the infrastructure needed to handle millions of streams simultaneously.

On the other hand, there are some sites which can release an anime fansub on a single server (1 server with a gigabit connection, seeding thousands of torrents stored on a SAN), and within an hour, well over 100,000 users can have a copy of that video even if it is a large 10GB+ 1080p 10 bit. One issue with services like YouTube, is that it is not profitable even for google, they instead use the platform to expand their audience in order to increase the value of the ads that they can put in other locations. This is all due to the internet not really being designed to function this way. It is far more efficient to download videos where if the segments of the download are downloaded in order, if it goes fast enough, you can download and watch at the same time, and if not, you wait for the download to finish (if the file support partial downloads). As long as a reasonable number of users upload as they download, the network scales well.

The benefit of bittorrent is that it requires surprisingly little infrastructure to host petabytes of content. With intelligent seeding where users with the best initial ratios get pieces first, you can often see a torrent with 1 seeder that was just released, along with 20,000 peers, still allow you to download at 100+Mbps, especially on networks where poor ratio peers cannot connect to seeders early on.
The other benefit of the network is that they are extremely receptive to a sudden influx of peers, as more join in, they have more throughput to share, and as the initial rush of viewers dies down, and the number of seeders drops, the few users that occasionally stop by to download the torrent, will easily be handled by the remaining seeders and the primary server. This is why some PC games use a torrent based updater, it allows an indie studio to push out a 500MB patch to a 100,000+ players, using a server on a gigabit connection.

The area where they can beat YouTube, is in the area of quality. For example, YouTube's compression is always a topic of complaint since even they need to keep the infrastructure for handling streaming under control. If torrent based distribution is done, someone could deliver a video that is 4K 200Mbps data rate. The major channels can have HD-8K+ streaming at similar quality to youtube, while smaller channels can just have standard definition with torrent based HD to 8k+ content. That will allow for far less distribution infrastructure as they will only really be dealing a small number of channels, while the vast majority will use torrent distribution. The infrastructure will just be storage heavy but with only a tiny fraction of the distribution infrastructure, thus reducing the cost of the most expensive aspect of the media platform.
« Last Edit: December 03, 2017, 01:33:11 pm by Razor512 »
 

Online coppice

  • Super Contributor
  • ***
  • Posts: 3678
  • Country: gb
Re: eevBLAB #41 - VidMe Is Shutting Down
« Reply #3 on: December 03, 2017, 01:46:39 pm »
Since YouTube is not profitable, a competitor that can actually be successful, will need to work towards a lower cost way to distribute the content.
YouTube has many faults, but their way of smoothly feeding video to millions spread far and wide has been pretty well honed. If they aren't profitable its a problem with their revenue, not their operating expenses. The question is just how far can they push their revenues? Clearly they would like to soak up a lot of the advertising cash currently going to broadcast TV. Their viability will depend on how far they get with that goal.
 

Offline Razor512

  • Regular Contributor
  • *
  • Posts: 132
  • Country: 00
Re: eevBLAB #41 - VidMe Is Shutting Down
« Reply #4 on: December 03, 2017, 02:14:53 pm »
They seem to use the platform more as a way to increase the user base of their ecosystem overall.
For example, if you have a channel with only 10,000 subscribers, an advertiser will not be willing you pay as much for an ad spot compared to someone with millions of subscribers.

This is why pay walls for news sites often cause them to lose even more money, they gain some direct income from a small number of readers, but they lose a significant amount of ad income since the target audience becomes smaller.

YouTube seems to me to be part of an overall ecosystem boost where the goal is not to have every aspect to be profitable, but instead use 1 aspect as an investment where other more profitable aspects of the business become even more profitable.

With youtube, the vast majority of videos, do not generate any income for the platform, but it does not stop a 10 subscriber channel from being able to upload a 10 hour long 8K video.
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 26841
  • Country: au
    • EEVblog
Re: eevBLAB #41 - VidMe Is Shutting Down
« Reply #5 on: December 03, 2017, 11:15:43 pm »
For example, if you have a channel with only 10,000 subscribers, an advertiser will not be willing you pay as much for an ad spot compared to someone with millions of subscribers.

Why would the CPM ad rate be any different?
If anything, CPM ad rates go down with increasing views.
 

Online coppice

  • Super Contributor
  • ***
  • Posts: 3678
  • Country: gb
Re: eevBLAB #41 - VidMe Is Shutting Down
« Reply #6 on: December 03, 2017, 11:20:16 pm »
For example, if you have a channel with only 10,000 subscribers, an advertiser will not be willing you pay as much for an ad spot compared to someone with millions of subscribers.
You do realise that this advertising is paid per view, don't you? This is not a broadcast medium. It doesn't matter if the view is part of only a 100 for that video or a billion, its still one person getting the ad shoved in their face.
 

Offline SparkyFX

  • Regular Contributor
  • *
  • Posts: 190
  • Country: de
Re: eevBLAB #41 - VidMe Is Shutting Down
« Reply #7 on: December 04, 2017, 07:34:33 pm »
Well, google at one point made the move to make every google account also a youtube account - without consent (not as far as i remember, it was surely covered as being a new google service). That´s not a big problem, you can log out and never use it. But sooner or later you stay logged in and watch and like and subscribe. That really got me, although i tried not to get into this video watch behavior algorithm thingy or needed to comment at all. But yeah, you find some interesting channels and want them to increase in their rating so you subscribe and like and comment, which requires/is more convenient by staying logged in. And without a (useful) feedback most people would not endure doing the huge effort to make it look appealing, so it would be a relatively short venture.

The move to connect google and youtube accounts suddenly made ads on youtube much more valuable than before, gave the platform a huge rise and the embedding in google search results made it even more attractive, because it links profiles with google mail, google plus and so on. I mean they track everything even without using an account by cookies, track pixels (IP Address based) and E-Tag cache ids, but with an associated account the data is worth much more, as suddenly age, gender and whatever other data is known and that even survives a cache cleanup - or is connected to a phone number and address. So all "Do not track" and anonymity aside, most people want to stay logged in anyway and thats in TANSTAAFL-Land ( not the harsh interpretation of it, i kind of interpret it with Open HW/SW in mind, as people still make a living with consultation, implementation and changes). Google dominates there, cause wherever you go they have it connected.
Working in electronics is like a good puzzle.
Comes with: More Parts! Complications™! Physics! English! Pulled out Hairs! Batteries included!
 

Online thm_w

  • Frequent Contributor
  • **
  • Posts: 772
  • Country: ca
Re: eevBLAB #41 - VidMe Is Shutting Down
« Reply #8 on: December 05, 2017, 10:22:55 am »
On the other hand, there are some sites which can release an anime fansub on a single server (1 server with a gigabit connection, seeding thousands of torrents stored on a SAN), and within an hour, well over 100,000 users can have a copy of that video even if it is a large 10GB+ 1080p 10 bit. One issue with services like YouTube, is that it is not profitable even for google, they instead use the platform to expand their audience in order to increase the value of the ads that they can put in other locations. This is all due to the internet not really being designed to function this way. It is far more efficient to download videos where if the segments of the download are downloaded in order, if it goes fast enough, you can download and watch at the same time, and if not, you wait for the download to finish (if the file support partial downloads). As long as a reasonable number of users upload as they download, the network scales well.

Seems like you've already realized the problem, there are no great P2P streaming services.
Sure you can click download and have the file within an hour, but, to click play and have the file streaming within a few seconds is a lot more difficult. You also need to store large amounts of data on the users device (tricky if you're on a phone).
It is probably solvable, I just haven't personally seen it done yet.
 

Offline Red Squirrel

  • Super Contributor
  • ***
  • Posts: 1732
  • Country: ca
Re: eevBLAB #41 - VidMe Is Shutting Down
« Reply #9 on: December 06, 2017, 02:09:37 pm »
Too bad, it would really be good to have a YT competitor, but it's very hard.  The way bandwidth is charged for is also very archaic so it makes it very hard for someone to just start their own data centre, because you get charged per usage and not just a single flat fee for a given size of pipe.  It's crossed my mind how it would be cool to start a small data centre but it's the internet connection that is the real crazy cost.  Not to mention most ISPs don't even allow to run servers let alone try to run something big like a video site.

If someone can get their hands on an unlimited gig connection with option to go higher then it would be a great start to starting such a service.  The key is to try to minimize monthly costs but bandwidth is just so expensive not to mention not always available.  In the places where it would be available it tends to be big cities where everything in general is going to be way more expensive such as property.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf