See http://www.tivo.com/ for the corporate website, http://www.avsforum.com/ubbcgitivo/Ultimate.cgi for TiVo community boards. The other big brand in the same market is Replay-TV at http://www.replaytv.com/.
I bought a Phillips DVR/Tivo unit in late 1999, and it was one of the best purchases I ever made. I don't watch a lot of TV, and the few shows that I like are often shown at inconvenient times. Using my VCR was simply too much effort for most shows, and I hated the noticably lower quality of the recorded version.
With the Tivo, my television habits changed drastically. I no longer care when shows are broadcast. Anything recorded is kept for at least two days (unless one requests to keep it longer), so I often watch shows from a few days ago. Television became more like a menu of selections (up to two weeks in the future) rather than a time-oriented performance medium.
Pausing and replaying live TV is incredibly neat--I often replay good action scenes in slow motion, or jump back to catch a bit of dialogue I missed. Advertisements are tiny 3-5 second blurs. Even when watching live TV, I often just press pause at the first commercial break, do something else for 10-20 minutes, then resume my slightly delayed viewing, jumping ahead at every commercial break.
The Tivo uses a general-purpose computer internally, and it uses the Linux operating system. There is no trace of the "computer" visible in operation, however. The system boots immediately to a graphical "please wait" screen rather than displaying diagnostics or booting progress. The unit is completely controlled through the remote--there isn't even an off button on the recorder. Aside from power and recording status lights, there are no other visible interfaces on the unit. The Tivo does take a bit of setup work, especially if you have a VCR and a satellite/cable box (which the Tivo can control through an included "IR blaster"), but it was mostly straightforward.
The Tivo unit I bought was the "14 hour" version, which stores about 7 hours of video at a "high" quality setting (better than VCR quality), or 4 hours at its best setting (similar to DVD quality). I usually can't tell the difference between "high" and "best" without still frames, so I usually use the "high" quality setting (which can be changed for individual recordings). Additionally, the Tivo continually records a separate 30 minute live-TV playback buffer at the best quality. There is now a "30 hour" unit which doubles the above times.
The TV scheduling information is downloaded nightly over a phone line, along with occasional software updates. Generic non-identifying viewing information is also uploaded, although it is possible to opt-out of that data collection. The scheduling service costs $10/month, although one can buy a lifetime subscription for about $200. (The Replay recorders bundle the lifetime service into their basic units, which cost about $200 more than the Tivo units.)
I highly recommend the Tivo recorder if you watch any significant amount of TV, especially if you find yourself watching advertisements. Just saving 10+ minutes/hour by skipping ads made the Tivo worthwhile for me. --CliffordAdams
One of the worries of many people was that DVRs would upset the current advertising-supported broadcast-TV systems. The Replay recorder includes a "30 second skip" which instantly (sub-second) skips exactly 30 seconds forward. The Tivo people have decided to work with the TV industry, and do not offer such a feature--instead they offer three fast-forward speeds (about 3x, 20x, and 60x speeds). (Both units have an "instant replay" feature which skips backwards about 10 seconds.)
The Tivo company has gained far more support from the TV industry due to their move. At least one person has claimed that the fast-forwarded versions may be even more effective than traditional advertisements, since people watch them closely to find the ending. (The 60X speed is too fast for most commercial breaks--one winds up spending more time rewinding after overshooting the end of the commercials.) Even the Replay company now claims that the 30 second skip is for "skipping slow scenes" rather than avoiding commercials.
The advertisement compromise is probably an indicator of future trends. It's actually better for the consumers than the trend in the DVD industry (which is almost totally "media-controlled"), with region-encoding and mandatory commercials (that disable the fast-forward feature).
I have one of these now. The fast-forward turns a 40 second advert into a 2 second advert with a "Tell me more" option for people who are interested. It's not too bad from an advertisers point of view. -- DaveHarris
The boundaries are blurred. It would make a lot of sense if the "2-second adverts" linked to web-based presentations instead of 40-second ads. TiVo has its own phone connection so is already just 3 hairs short of being on the Internet. Advertisers could well want downloadable executable content such as Java applets and JavaScript.
Also, you can get a shell prompt out of a TiVo. Follow the links at the top of this page to find a community of TiVo hackers. -- DaveHarris
TiVo is very much like having your own personal TV channel, whose content is taken from the best content of all the other channels, and which broadcasts to a schedule of your choosing. Although it isn't implemented as a broadband internet appliance, it feels like one. It uses broadcast plus local caching to simulate narrowcast.
Let's talk about caching. My TiVo stores 20 hours of video on a 40 Gig disk. Soon (5 years?) terabyte drives will be cheap. When you can store 500 hours, you don't care about downloading stuff you will never use. You could store a substantial fraction of the WWW. Do we really need ISDN when everything is local? With caching we lose some interactivity, but some people like it that way. People talk of TvComputerConvergence? as being more interactive than normal TV but it may also be less interactive than the web usually is.
TiVo acts as a SoftwareAgent? that seeks out useful content. I think it is very like ViewPoint in the way it collects and assembles information. There is a connection here with the BookMetaphor. A book is one form of organised, high-quality presentation. ViewPoint is partly about finding a human editor, or SoftwareAgent?, who can impose that organisation on the raw material. Interaction would just get in the way.
The ideal advert would last a couple of seconds and basically say, "Fancy a pizza?" If you don't, ignore it and it'll fade. If you do, click and get a 40 second presentation and a menu. Make your choice, and TiVo dials out for you. TiVo doesn't have an Internet connection, but it has a phoneline and it knows your postcode so it can find a local vendor who does free delivery. Pay the pizza boy when he comes as normal. This doesn't need ISDN, doesn't need a lot of interactivity. All the video content is served locally, having been downloaded in the wee hours a week before. The advertisers will be renting at least half of that Terabyte disk.
Firstly, a 2-second ad will help to establish a brand name. Secondly, I agree that the new and the old tend to coexist rather than the new totally displace the old. In this way does modern life become more complex. Thirdly, even if it goes against the advertisers interests, I don't see why the advertisers will get their way.
I see the 2-second ad (with "tell me more" link) as a feasible compromise. If ads are more intrusive than that, people will seek out ways to get around them. Fast-forward buttons, software that strips out banner ads etc. A TiVo-like box that forces you to watch intrusive adverts will ultimately fail in the market place, if it is possible to make a TiVo-like box which doesn't. And I think it will be possible; I don't think advertisers can hold back the tide here. The issues are similar to those of copyright and Napster (cf NapsterDotCom, InformationWantsToBeFree).
One of my old projects was the "strn" UseNet newsreader, which added many new features to the trn newsreader. One of these features was "Virtual Newsgroups", which allowed people to create and even share collections of articles across multiple newsgroups. Users could select individual articles or the highest-scoring articles from newsgroups. In some newsgroups I would read only a few people's contributions, like Henry Spencer's articles in the space newsgroups. One of my favorite practices was to make a huge virtual newsgroup of all my favorite groups, and read only the top-scoring articles. Unfortunately, this often took far too much CPU/memory/IO for the machines of that time (when a gigabyte was a lot of disk space), so I only did it rarely. --CliffordAdams