Quote: Original post by Eelco
A friend of mine is researching protocols for on-demand television. He told me about the politics of it once; im paraphrasing, but it has a lot to do with investments that people arnt quite mentally ready to declare sunk yet. Its an industry that hasnt seen any real change for deccades, run by complacent morons. They are starting to realize they will have to catch up sooner or later or bite the dust though. Give it a few more years.

The good news for those people is that all their trillion-dollar hard investments (cable, satellite, and telecom infrastructure) are perfectly good for carrying TCP/IP traffic, and in fact are already being widely used to that effect.

Quote: Original post by BeanDog

Quote: Original post by Eelco
A friend of mine is researching protocols for on-demand television. He told me about the politics of it once; im paraphrasing, but it has a lot to do with investments that people arnt quite mentally ready to declare sunk yet. Its an industry that hasnt seen any real change for deccades, run by complacent morons. They are starting to realize they will have to catch up sooner or later or bite the dust though. Give it a few more years.

The good news for those people is that all their trillion-dollar hard investments (cable, satellite, and telecom infrastructure) are perfectly good for carrying TCP/IP traffic, and in fact are already being widely used to that effect.

Not sure about that. You can not assume any coherence between what people are watching. Current television infrastructure is calculated to carry a few hundered of television-quality signals. If you are aiming for real-time on demand viewing, you are either going to have to place a distribution hub on every street, or radically scale bandwidth. Either way, its going to require infrastructural change of some sort. Which has to be coordinated with hardware manufacturers. Not that it couldnt have happend five years ago, but its not trivial either.

Personally, im perfectly happy just to ditch this whole television thing altogether.

Quote: Original post by Eelco

Quote: Original post by BeanDog

Quote: Original post by Eelco
A friend of mine is researching protocols for on-demand television. He told me about the politics of it once; im paraphrasing, but it has a lot to do with investments that people arnt quite mentally ready to declare sunk yet. Its an industry that hasnt seen any real change for deccades, run by complacent morons. They are starting to realize they will have to catch up sooner or later or bite the dust though. Give it a few more years.

The good news for those people is that all their trillion-dollar hard investments (cable, satellite, and telecom infrastructure) are perfectly good for carrying TCP/IP traffic, and in fact are already being widely used to that effect.

Not sure about that. You can not assume any coherence between what people are watching. Current television infrastructure is calculated to carry a few hundered of television-quality signals. If you are aiming for real-time on demand viewing, you are either going to have to place a distribution hub on every street, or radically scale bandwidth. Either way, its going to require infrastructural change of some sort. Which has to be coordinated with hardware manufacturers. Not that it couldnt have happend five years ago, but its not trivial either.

Personally, im perfectly happy just to ditch this whole television thing altogether.

Part of it is looking at how you can actually broadcast the signal in an online environment. Because that is partly what you still want to be doing. I don't really think we'll see full "on demand" that starts a few ms after you click it, and lets you jump around the full video at will. Or atleast we won't be seeing it anytime soon.

A better model would be modified broadcast. For a popular show, chances are that with in a minute more than one person is going to want to start watching it. So rather than streaming a copy to each and every person that requests it, the very moment they request it, have them wait 5 seconds for extremely popular shows before jumping on the next stream. Less popular, older, or more obscure clips of any length may have you waiting a minute with a few ads for other things you may be interested in viewing.

And also a software system to allow you to subscribe to given 'channels/programs', and will download a copy onto your local system on the first Broadcast stream.

It is one aspect of networking I never really understood. There are so many applications that are One to Many signaling, and yet from my understanding of modern networking (not a field I focus on) is that it is all based on One to One signals. If you want to send the same data to 3 different computers, you send address three boxes, throw the data in, and kick them onto the net. We are apparently suppose to be getting the ability to get a box that we can put data in and stick multiple address on, and have the box self replicate on its way to each of those addresses, splitting itself and keeping only its own part of the addresses going in its direction. With this we can see far better usage of "Broadcast" internet usage, and could really change how things work for streaming video, or large software updates going to millions of users. Think of how easy game updates would be then if the host servers could send out 1 data packet, and it is read by 1000 users, rather than a single packet to each user.

Quote: Original post by BeanDog
<good stuff>

I wrote that paper in 2005, to graduate. It will come. See below, for instance.

Quote: Original post by Codeka

Quote: Original post by borngamer
I also don't want to sit and wait for a show to download so a torrent is out of the question.

You don't "sit and wait" for it to download. You choose which shows you want to watch, and it downloads them in the background. Then, when you're ready, you watch whatever you want.

Actually, it streams. My Time Warner Cable digital television service is already streamed. I know this because Time Warner introduced a feature a few months ago called "Start Over," where if you join your show at any point in the middle, you can press the select button to bring up a menu and start over from the beginning. You can't scrub through a show that's been started over (which makes sense, since it's streaming, and you don't have a precise scrubber).

Quote: Original post by ChaosEngine
Ad-supported is not the way of the future, IMHO. I would happily pay a small per-episode fee to watch some shows uninterrupted. Maybe with an ad-supported, lower quality free version too, so you can check out a show before you watch it.

I also wrote this in my college paper. Essentially, the model would be that advertisers would pay you to watch their ads. How? By defraying the cost of your access. If you indicate that you're interested in specific product categories and willing to receive messages, you can receive a reduction in your cable access fee. On the other hand, if you would prefer not to receive advertising messages, then you get no access subsidies, but your show is uninterrupted.

(Did you see what I did there, too? Instead of making "ad-free" a premium, I made ad-supported a discount! Oh, Madison Avenue loves me. [smile])

Quote: Original post by BeanDog

Quote: Original post by LessBread
Is the advertising supported business model working out for content driven internet sites in general? If the model works for the most popular sites, is that sufficient for the less popular sites?

I recently read that YouTube is bleeding a few million bucks a day, so maybe not.

While true of YouTube, note that YouTube is in the unpalatable position (to advertisers) of trying to place messages next to arbitrary user-generated content. Hulu is making money hand over fist.

Quote: Original post by LessBread
That sounds good to me, but I'm not in the advertising business. How would a television station sell advertising if it couldn't guarantee the time let alone the viewers?

Television stations are dead. It will be the cable provider selling the advertising. Content production studios - which is what "television stations" will become - will produce newscasts, shows, etc, and the cable access provider will sell advertising around them, remitting a portion of the proceeds to the producer of the individual show. Advertising will be sold on a much more targeted basis rather than in aggregate based on projections - Nielsen is fucked, too; the cable providers will have all the audience demographic and viewing habit data they need, almost real time.

The one great loss in this model is "live events" (though not totally; it can be streamed, and users can elect to join the stream or even start over at its beginning) or simultaneous viewing. Remember the Seinfeld finale, which virtually the whole country watched at the same time? Yeah. That's never happening again.

Quote: Original post by ChurchSkiz
I have two real-world experiences with which to back up my position. I think it's a good idea but I certainly can see why it's not happening on a large scale.

1. I had network tv (no cable), for 3 years. We watched what was on tv and were ok with it. After purchasing cable, I feel like nothing good is ever on and now spend my time "surfing" instead of forcing myself to watch a show that is not my top pick out of necessity.

Given a broad choice of what to "watch", I can see people spending more time looking at choices than watching shows, which translates to less time watching ads.

Have you ever used a cable on-demand system? Your model and notion of television is purely synchronous, so you haven't quite considered usage habits and possibilities in an asynchronous model. When you turn to one of your "On Demand" channels, a rotating video of other on-demand programming - advertising, essentially - starts playing. You browse using about 75% of the screen, while the ad remains in place until you begin playing your selection. It's not too hard to imagine that looking for shows - "surfing" - will be accompanied by a rotating selection of messages from any advertisers you have indicated interest in or willingness to accept messages from.

Secondly, the reason you channel surf is because you have such poor information about what's on now. But in an asynchronous, internet-based model, a.) "now" is an irrelevant concept, as you can pull up anything at any time, and b.) the system can notify you of new episodes of content you're interested in, or of content that is similar - it can even crowdsource the similarity using a system akin to Amazon's "47% of users who bought X also bought Y".

Quote: 2. Psychologically something happens when you have the ability to watch something at any given time. The best example of this that I have is my DVD collection. I love The Matrix, it is one of my favorite movies, but I have only watched my DVD version once. Whenever I am about to pop it in I say, "Nah, too many boring parts." However, without fail, if I see it on tv I will watch the entire movie, commercials and all. This happens with almost all of my dvd's.

That's re-watching the same old content. Do you have DVR/TiVo? Do you skip new episodes of shows that you've recorded because there might be "too many boring parts"?

I didn't think so.

Quote: Original post by Oluseyi
Hulu is making money hand over fist.

How much is that? Can you put a number to that or is that just a feeling you have?

Quote: Original post by Oluseyi

Quote: Original post by LessBread
That sounds good to me, but I'm not in the advertising business. How would a television station sell advertising if it couldn't guarantee the time let alone the viewers?

Television stations are dead. It will be the cable provider selling the advertising. Content production studios - which is what "television stations" will become - will produce newscasts, shows, etc, and the cable access provider will sell advertising around them, remitting a portion of the proceeds to the producer of the individual show. Advertising will be sold on a much more targeted basis rather than in aggregate based on projections - Nielsen is fucked, too; the cable providers will have all the audience demographic and viewing habit data they need, almost real time.

So you're saying that television stations won't sell advertising because they can't guarantee the time or the viewers and that even though they're dead they'll be able to generate content? It's hard to tell, because you seem to be answering a different question than the one I posed.

Quote: Original post by Oluseyi
The one great loss in this model is "live events" (though not totally; it can be streamed, and users can elect to join the stream or even start over at its beginning) or simultaneous viewing. Remember the Seinfeld finale, which virtually the whole country watched at the same time? Yeah. That's never happening again.

I was not a Seinfeld fan, so I didn't see the finale. I did watch the finales of Cheers and MASH. I've also watched the Super Bowl every year since, gosh, 1976? Let's just say for a really long time. It seems to me that the end of "live events" presents a terrible problem for televised sports.

It also seems to me that the whole country watches national tragedies at the same time - space shuttle disasters, 9/11, Katrina etc.

Quote: Original post by Oluseyi
Secondly, the reason you channel surf is because you have such poor information about what's on now. But in an asynchronous, internet-based model, a.) "now" is an irrelevant concept, as you can pull up anything at any time, and b.) the system can notify you of new episodes of content you're interested in, or of content that is similar - it can even crowdsource the similarity using a system akin to Amazon's "47% of users who bought X also bought Y".

I disagree with your reasoning here. I channel surf because I don't want to watch commercials. I suppose, however, that when I know what's on now, I don't surf, so much as I switch back and forth between shows hoping they don't have their commercials synchronized. I should say, however, that channel surfing is positive if it leads people to watch shows they would not ordinarily watch. In that, it has the potential to broaden horizons. It's difficult to know if you're interested in something if you haven't seen it before. This points out a downside to content immediacy. It easily leads to narrow tastes. And when it comes to "crowdsourcing", I've always been skeptical of it as a marketing ploy, so I generally ignore it.

Quote: Original post by LessBread

Quote: Original post by Oluseyi
Hulu is making money hand over fist.

How much is that? Can you put a number to that or is that just a feeling you have?

They don't release numbers, but they're doing quite well. I've only heard whispered figures.

Hulu doesn't have the bandwidth requirements that YouTube does because it hosts less content - only professional content from its partner sites, as opposed to the "all the world's burping baby videos" strategy of YouTube. This means lower cost. Then Hulu can place ads on every single video it has, because advertisers have guarantees about the nature of the content - something YouTube can't do ever (the hope of being able to algorithmically match advertising messages to videos scares advertisers; what if your ad ends up next to something offensive or controversial?), and on longer content it actually places multiple ads. So the combination of lower overhead, greater inventory and better rates means of all the video sites, Hulu is making the most right now.

Quote: So you're saying that television stations won't sell advertising because they can't guarantee the time or the viewers and that even though they're dead they'll be able to generate content? It's hard to tell, because you seem to be answering a different question than the one I posed.

I'm saying television stations won't sell advertising because television stations won't exist. If your content is entirely on-demand via an internet connection, why are you switching to channel 3 again? What does "channel 3" even mean in that context?

No, television stations don't become sites. Nobody wants to have to browse different sites or discover and subscribe when they're sitting back at the television. It's just a bad user experience. The cable operator becomes the sole site, managing your subscriptions and preferences and funneling you recommendations. IP TV is individual shows, regardless of the originating production body. You'll be able to queue up CNN, FOX News and ABC World News as a block of programming that you watch when you get home, if you want, rather than channel surfing between the three at the same time. Television stations become production houses, and instead of spending money operating broadcast stations they spend their money producing higher quality content in order to drive traffic and a better revenue split with the cable operator.

Quote: I was not a Seinfeld fan, so I didn't see the finale. I did watch the finales of Cheers and MASH. I've also watched the Super Bowl every year since, gosh, 1976? Let's just say for a really long time. It seems to me that the end of "live events" presents a terrible problem for televised sports.

It also seems to me that the whole country watches national tragedies at the same time - space shuttle disasters, 9/11, Katrina etc.

And they will. While the bulk of viewing will likely be pre-recorded, asynchronous content, IP technology allows for live streaming - like all the sites that let us watch the Inauguration on our computers this year, for instance. (Incidentally, I most enjoyed FOX News' coverage on Hulu.)

Say you're watching some gripping History Channel documentary and a disaster occurs. All news-gathering organizations will immediately begin publishing live feeds with coverage. If you have a priority list of news sources set up, you will be notified of their availability in cascading order from highest rated to lowest. If not, you may receive multiple notifications that "Such and such has just happened! Press SEL to learn more."

You press the SEL button. Your current stream is stopped, with your position marked. The live stream corresponding to the source you've chosen loads. Voilà, live coverage.

(Alternately, your stream may not be stopped immediately, but instead shrunk to 25% of the screen while information about the event and available streams occupies the other 75%, allowing you to decide whether you wish to switch streams and which stream you wish to switch to.)

Quote: I disagree with your reasoning here. I channel surf because I don't want to watch commercials. I suppose, however, that when I know what's on now, I don't surf, so much as I switch back and forth between shows hoping they don't have their commercials synchronized.

So earlier we discussed the fact that you could have absolutely no commercials in your stream, or only select commercials - and you could possibly select how frequently you wish to see commercials, on a tiered basis, in return for paying less. When avoiding commercials is no longer a concern, and you don't have to switch back and forth between shows because missing shows is a thing of the past, since you can just queue it up when you're ready, why would you channel surf?

Quote: I should say, however, that channel surfing is positive if it leads people to watch shows they would not ordinarily watch. In that, it has the potential to broaden horizons. It's difficult to know if you're interested in something if you haven't seen it before. This points out a downside to content immediacy. It easily leads to narrow tastes. And when it comes to "crowdsourcing", I've always been skeptical of it as a marketing ploy, so I generally ignore it.

"Crowdsourcing" isn't about marketing. It's about capturing closer approximations of general trends and tendencies. That approximation is generally used to drive marketing, but in a fashion that isn't about a content publisher enticing you to consume a specific product but rather an intermediary suggesting additional content you might be interested in. It's YouTube "Related Videos" vs "Coming Up Next..."

Instead of relying on a station programmer who says, "People who watch this station like these shows. Nielsen tells us so!", the recommendation engine can drill down to as obscure a datum as it can find, and employ as many variables as it can manage to cross-index stuff you've watched, stuff you've subscribed to, and stuff that other people have watched and subscribed to to suggest other stuff you might want to watch or subscribe to.

But, hey, even if you "miss" it - which is an anachronism, since it doesn't play until you queue it up, ever - you just need to hear about the show via word of mouth and then you can watch it from episode 1, whenever you're ready.

So many of the base assumptions about television viewing habits are rooted in the fact that it's been a synchronous activity with an opportunity cost of "missing something." Take that out of the equation and habits change entirely.

Quote: Original post by Oluseyi

Quote: Original post by LessBread

Quote: Original post by Oluseyi
Hulu is making money hand over fist.

How much is that? Can you put a number to that or is that just a feeling you have?

They don't release numbers, but they're doing quite well. I've only heard whispered figures.

So you can't put any numbers to it?

Quote: Original post by Oluseyi
Hulu doesn't have the bandwidth requirements that YouTube does because it hosts less content - only professional content from its partner sites, as opposed to the "all the world's burping baby videos" strategy of YouTube. This means lower cost. Then Hulu can place ads on every single video it has, because advertisers have guarantees about the nature of the content - something YouTube can't do ever (the hope of being able to algorithmically match advertising messages to videos scares advertisers; what if your ad ends up next to something offensive or controversial?), and on longer content it actually places multiple ads. So the combination of lower overhead, greater inventory and better rates means of all the video sites, Hulu is making the most right now.

That explains how Hulu could be profitable compared to YouTube, but it still leaves open just how profitable Hulu is so far.

Quote: Original post by Oluseyi

Quote: So you're saying that television stations won't sell advertising because they can't guarantee the time or the viewers and that even though they're dead they'll be able to generate content? It's hard to tell, because you seem to be answering a different question than the one I posed.

I'm saying television stations won't sell advertising because television stations won't exist. If your content is entirely on-demand via an internet connection, why are you switching to channel 3 again? What does "channel 3" even mean in that context?

In that context Channel 3 becomes an ip address providing content that differs from other ip addresses. In other words, instead of there being one point for on-demand content, there are millions.

Quote: Original post by Oluseyi
No, television stations don't become sites. Nobody wants to have to browse different sites or discover and subscribe when they're sitting back at the television. It's just a bad user experience. The cable operator becomes the sole site, managing your subscriptions and preferences and funneling you recommendations. IP TV is individual shows, regardless of the originating production body. You'll be able to queue up CNN, FOX News and ABC World News as a block of programming that you watch when you get home, if you want, rather than channel surfing between the three at the same time. Television stations become production houses, and instead of spending money operating broadcast stations they spend their money producing higher quality content in order to drive traffic and a better revenue split with the cable operator.

I think they ought to become sites, otherwise, we're looking at a reduction of content availability even as the technology provides for an explosion of content. Why limit yourself to one on-demand source provided by the cable company when there could be thousands if not millions of them? Channels in this model become clearinghouses just like on the internet. And just like on the internet, viewers frequent those sites they find the most useful. Cable news at present doesn't broadcast (along the lines of the model your setting forward), yet the quality of content has declined. News has become infected with entertainment in order to attract viewers. That only gets worse under the model you're describing. Quality doesn't drive news traffic, at least not under the current news business model. To suggest that will change under this new model seems to me like wishful thinking. Worse, this model looks to me like the start of a monopoly, the very thing that has the net neutrality activists worried.

Quote: Original post by Oluseyi

Quote: I was not a Seinfeld fan, so I didn't see the finale. I did watch the finales of Cheers and MASH. I've also watched the Super Bowl every year since, gosh, 1976? Let's just say for a really long time. It seems to me that the end of "live events" presents a terrible problem for televised sports.

It also seems to me that the whole country watches national tragedies at the same time - space shuttle disasters, 9/11, Katrina etc.

And they will. While the bulk of viewing will likely be pre-recorded, asynchronous content, IP technology allows for live streaming - like all the sites that let us watch the Inauguration on our computers this year, for instance. (Incidentally, I most enjoyed FOX News' coverage on Hulu.)

I enjoyed the free concert on HBO. Bruce Springsteen and Pete Seeger singing "This Land is Your Land" with a gospel choir on the steps of the Lincoln Memorial. Wow!

Quote: Original post by Oluseyi
Say you're watching some gripping History Channel documentary and a disaster occurs. All news-gathering organizations will immediately begin publishing live feeds with coverage. If you have a priority list of news sources set up, you will be notified of their availability in cascading order from highest rated to lowest. If not, you may receive multiple notifications that "Such and such has just happened! Press SEL to learn more."

You press the SEL button. Your current stream is stopped, with your position marked. The live stream corresponding to the source you've chosen loads. Voilà, live coverage.

(Alternately, your stream may not be stopped immediately, but instead shrunk to 25% of the screen while information about the event and available streams occupies the other 75%, allowing you to decide whether you wish to switch streams and which stream you wish to switch to.)

Sounds to me like a more likely use of that feature would be to spam me with commercials for crap I don't want at exactly the moments I don't want to be spammed. It's bad enough that television has started to plaster promos into the lower screen during shows or that television never interrupts a commercial for an emergency alert system test or an amber alert. If I sound cynical about this, it's because I am. I'm old enough to remember the lies used to sell cable television 30 years ago.

Quote: Original post by Oluseyi

Quote: I disagree with your reasoning here. I channel surf because I don't want to watch commercials. I suppose, however, that when I know what's on now, I don't surf, so much as I switch back and forth between shows hoping they don't have their commercials synchronized.

So earlier we discussed the fact that you could have absolutely no commercials in your stream, or only select commercials - and you could possibly select how frequently you wish to see commercials, on a tiered basis, in return for paying less. When avoiding commercials is no longer a concern, and you don't have to switch back and forth between shows because missing shows is a thing of the past, since you can just queue it up when you're ready, why would you channel surf?

To avoid commercials. I don't believe promises that there could be "absolutely no commercials" in the stream. That was how cable television was sold 30 years ago. Today they stick commercials into the middle of on-demand shows. I fast forward through them. And even though the technology is supposed to be non-linear, they don't provide a content bar that would allow me to instantaneously jump over the commercials. It's a game of cat and mouse ultimately.

Quote: Original post by Oluseyi

Quote: I should say, however, that channel surfing is positive if it leads people to watch shows they would not ordinarily watch. In that, it has the potential to broaden horizons. It's difficult to know if you're interested in something if you haven't seen it before. This points out a downside to content immediacy. It easily leads to narrow tastes. And when it comes to "crowdsourcing", I've always been skeptical of it as a marketing ploy, so I generally ignore it.

"Crowdsourcing" isn't about marketing. It's about capturing closer approximations of general trends and tendencies. That approximation is generally used to drive marketing, but in a fashion that isn't about a content publisher enticing you to consume a specific product but rather an intermediary suggesting additional content you might be interested in. It's YouTube "Related Videos" vs "Coming Up Next..."

Isn't about marketing? We're talking about a book seller not a library. It's all about marketing, about selling more product. The difference is that by providing this service the marketer seeks to build trust and brand loyalty. The net allows the marketer to mask that service behind other customers, who we're supposed to see as disinterested, that is, as not driven by profit. The very language to describe these services drips with marketing. "Crowdsourcing" sounds like something drummed up by someone making a sales pitch.

Quote: Original post by Oluseyi
Instead of relying on a station programmer who says, "People who watch this station like these shows. Nielsen tells us so!", the recommendation engine can drill down to as obscure a datum as it can find, and employ as many variables as it can manage to cross-index stuff you've watched, stuff you've subscribed to, and stuff that other people have watched and subscribed to to suggest other stuff you might want to watch or subscribe to.

Sounds like total information awareness with a profit motive. Are we supposed to embrace our categorization? Is being stuffed into a box supposed to make our lives easier? We know you'll like this, the mathematics proves it!

Quote: Original post by Oluseyi
But, hey, even if you "miss" it - which is an anachronism, since it doesn't play until you queue it up, ever - you just need to hear about the show via word of mouth and then you can watch it from episode 1, whenever you're ready.

What worth is there to word of mouth when everything is indexed and cross referenced?

Quote: Original post by Oluseyi
So many of the base assumptions about television viewing habits are rooted in the fact that it's been a synchronous activity with an opportunity cost of "missing something." Take that out of the equation and habits change entirely.

That has been used as a chain to yank people around. They'll figure out a new chain to replace the old one.

Sorry if I don't share your enthusiasm.

Quote: Original post by LessBread
Sorry if I don't share your enthusiasm.

*shrug* Sorry if I don't share your boundless pessimism.