IP TV - TCP-IP

This is a discussion on IP TV - TCP-IP ; On Jan 1, 5:08*pm, Ethan Howe wrote: > For example, the iPod > vs. no-name, low-cost MP3 players from Asia. Or the aforementioned Xbox > against the PC. The fact that a PC can probably be configured to play > ...

+ Reply to Thread
Page 2 of 2 FirstFirst 1 2
Results 21 to 25 of 25

Thread: IP TV

  1. Re: IP TV

    On Jan 1, 5:08*pm, Ethan Howe wrote:

    > For example, the iPod
    > vs. no-name, low-cost MP3 players from Asia. Or the aforementioned Xbox
    > against the PC. The fact that a PC can probably be configured to play
    > games does not make it automatically the easiest and most desirable
    > solution for the general public. In fact, a startup company a few years
    > ago tried to sell a set-top video game console that did not try to hide
    > the fact that it was a commodity PC using off-the-shelf parts and open
    > source operating system and developer tools. Despite the obvious virtue
    > in such an idea, the product never materialized despite millions of
    > investor dollars being thrown at the idea.


    I think the average joe out there likes a simple little toy he can
    easily navigate, the type of toy Apple excells in developing, and that
    such simplicity becomes increasingly difficult as you add on
    multipurpose functions to the device. That, I think, is the root of
    the problem. A PC will inherently lose out to purpose-built, small,
    cute little toys that are optimized for their roles.

    I'd say the iPhone is perhaps the device to push those boundaries,
    even though of course it doesn't come close to doing everything a PC
    can do quite well.

    > While the jury is still out on IPTV, I believe that the same result is
    > possible: monopolistic, proprietary devices, if done well and packaged
    > right will succeed over the general purpose hardware.


    Since cable and DBS set-top boxes are also proprietary devices, and
    people seem eager enough to become dependent on those, I don't see why
    IPTV can't work. You just buy it as if you were buying cable TV, only
    the protocol is different. The average joe who wants whatever it takes
    to watch his sports shows probably couldn't care less.

    The marketers capitalized on the letters "IP" in IPTV, to hype up
    their offering. The innocent trade press very often parrotted their
    words without informing themselves. But as a TV distribution service,
    if priced right, I don;t see why IPTV can't be completely competitive
    with the other walled gardens. I don't think there are very daunting
    technical problems here, as long as the scope of the multicasts is
    limited.

    Bert

  2. Re: IP TV

    On Jan 1, 4:08*pm, Ethan Howe wrote:
    > While the jury is still out on IPTV, I believe that the same result is
    > possible: monopolistic, proprietary devices, if done well and packaged
    > right will succeed over the general purpose hardware.
    >
    > > I have a $180 million-dollar story about perception:

    >
    > [--snip--]
    >
    > > But it was too late. It was too late because fiction superseded fact.

    >
    > Wait, why did the company fail? Because the product had technical issues
    > that the engineers could not solve? The way you tell the story, you make
    > it sound like the the company failed because the marketing department
    > realized too late that they could have told the engineering department
    > to use commodity hardware? This does not make any sense: Since when does
    > the market department make technical recommendations to the engineering
    > department? And why would the engineers listen to a bunch of
    > salespeople? They did not listen to you, and you were hired as an
    > engineer, right?


    The company failed for multiple reasons, the most technical of which
    is that the product did not live up to customer expectations. It was
    a web browser. Customers would try to use it to browse, and it would
    choke on a large fraction of web pages because the software was
    proprietary almost from the ground up. The marketing people had to to
    start promoting the browser as a browser that could only work with a
    few basic graphic formats. No audio. Etc.

    It was not the sales and marketing departments' decision to make
    technical recommendations to the engineering people, but given that
    the engineering people were being unsensible (IMO), it might have
    helped if all parties with a stake in companies success were aware of
    all available options. I *think* the engineering people left the
    marketing and salespeople in the dark purposely to preempt any
    discussion of going with a commoditized OS and web browser.

    In any case, the problems were never fixed. When customers asked why
    they could not hear sound on the web pages, the answer was a shoulder
    shrug.

    > If the engineering department chose to go the proprietary path, then
    > they (you) should have been able to fix the problems with the LEDs
    > because it was truly proprietary and designed by your engineers. If the
    > engineering department really instead chose the commodity path
    > (regardless of whether they called it proprietary out of pride), so
    > what? They (you) should have been able to fix the problem more easily,
    > because you were using "a standard laptop computer straight out of
    > China." Either way, at the end of the day, the buck has to stop at the
    > engineering department for unsolved technical problems. It is not
    > responsibility on the marketing and sales department.


    The LED problem was trivial to fix. The other problems were not
    fixable given the path that had been chosen. We had more duct tape on
    that product than a 3M factory. Every time it booted without
    crashing, there was a sigh of relief on my part.

    -Le Chaud Lapin-

  3. Re: IP TV

    On Jan 1, 4:19*pm, Albert Manfredi wrote:
    > > > Le Chaud Lapin writes:

    >
    > > > > If there is any doubt in anyone's mind about #1, I challenge you to
    > > > > write a a multicast application for 1 million nodes in < 1 week that
    > > > > will work without heavy hand-holding and duct tape. IMO, if you
    > > > > cannot, and you are a skilled coder and relatively bright, and have a
    > > > > solid understanding of the principle of multicast, something is wrong
    > > > > with the model.

    >
    > The model works quite well. The problem is that most likely, a single
    > "relatively bright" coder doesn't have access to all the points in the
    > network that he needs to have access to, to enable multicast. So
    > whatever he does may not travel very far.


    Well, that kind of substantiates what I have been saying, don't you
    think? The whole point of having a multicast model is that the
    "relatively bright" coder should not have to have access to all the
    points in the network. S/he should be able to write the application,
    and it should work over the existing model.

    Actually this is interesting because, long ago before the age of
    laissez-faire packet switching, there were networking people who were
    disturbed at the idea that their packets would enter a network segment
    not under their control and magically arrive at its destination. They
    felt that it was more prudent that they controled everything, like bit-
    error rate, variance, security, reliability, etc. Today we take the
    routing infrastructure for granted.

    The same cannot be said for multicasting. One cannot say, "I think I
    will write a 'multicast' application for a target of 1,000,000 nodes
    on the Internet and deploy a prototype 6 months from now." Whatever
    application is written, when it "deploys", it will sit like a brick
    because, as you hinted, the infrastructure is not there. The model is
    broken.

    Note that by "model", I mean what one reads about in current RFC's,
    the implementations of multicast as it exists today, in 2007, not
    vague principles that are obviously true and might be properly
    implemented in 2020.

    -Le Chaud Lapin-

  4. Re: IP TV


    "Le Chaud Lapin" wrote in message
    news:fd97f4fb-9d9f-42dd-a7cb-e33217b76eef@k2g2000hse.googlegroups.com...
    On Jan 1, 4:19 pm, Albert Manfredi wrote:
    > > > Le Chaud Lapin writes:

    >
    > > > > If there is any doubt in anyone's mind about #1, I challenge you to
    > > > > write a a multicast application for 1 million nodes in < 1 week that
    > > > > will work without heavy hand-holding and duct tape. IMO, if you
    > > > > cannot, and you are a skilled coder and relatively bright, and have
    > > > > a
    > > > > solid understanding of the principle of multicast, something is
    > > > > wrong
    > > > > with the model.

    >
    > The model works quite well. The problem is that most likely, a single
    > "relatively bright" coder doesn't have access to all the points in the
    > network that he needs to have access to, to enable multicast. So
    > whatever he does may not travel very far.


    Well, that kind of substantiates what I have been saying, don't you
    think? The whole point of having a multicast model is that the
    "relatively bright" coder should not have to have access to all the
    points in the network. S/he should be able to write the application,
    and it should work over the existing model.

    Actually this is interesting because, long ago before the age of
    laissez-faire packet switching, there were networking people who were
    disturbed at the idea that their packets would enter a network segment
    not under their control and magically arrive at its destination. They
    felt that it was more prudent that they controled everything, like bit-
    error rate, variance, security, reliability, etc. Today we take the
    routing infrastructure for granted.

    The same cannot be said for multicasting. One cannot say, "I think I
    will write a 'multicast' application for a target of 1,000,000 nodes
    on the Internet and deploy a prototype 6 months from now." Whatever
    application is written, when it "deploys", it will sit like a brick
    because, as you hinted, the infrastructure is not there. The model is
    broken.

    Note that by "model", I mean what one reads about in current RFC's,
    the implementations of multicast as it exists today, in 2007, not
    vague principles that are obviously true and might be properly
    implemented in 2020.

    -Le Chaud Lapin-


    Thank you all for replying to my post re IPTV.
    One more thing,
    if I were offered a job as an IPTV installer from a Telco, should I take it?
    Let's say that the pay and the benefits are ok.

    thanks!



  5. Re: IP TV

    eager wrote:
    > "Le Chaud Lapin" wrote in message
    > news:4f90b54b-3d0f-429e-8c2d-c8fcaa71af99@e23g2000prf.googlegroups.com...
    > On Jan 1, 7:21 am, Bruce Barnett
    > wrote:
    >> Le Chaud Lapin writes:
    >>
    >>> If there is any doubt in anyone's mind about #1, I challenge you to
    >>> write a a multicast application for 1 million nodes in < 1 week that
    >>> will work without heavy hand-holding and duct tape. IMO, if you
    >>> cannot, and you are a skilled coder and relatively bright, and have
    >>> a solid understanding of the principle of multicast, something is
    >>> wrong with the model.

    >>
    >> The problem with multicast, as I see it, is a chicken and egg. It's
    >> the routers - and I mean all of the ones between the sender and the
    >> receivers - including firewalls - that have to support the multicast
    >> (IGMP) protocol.

    >
    > Yes. But this is a technical issue, IMO. The average user doesnt'
    > know, doesn't care. The do want multicast, though they not phrase it
    > that way when they ask for it:
    >
    > I spoke to a user that has an IPTV and he does not like it.
    > He said that ISP gives him 7 Mbps of bandwidth, 2 Megs for each TV, +
    > overheads ....
    > Basically, it sucks.


    You think that's bad. I know someone who's provider give him only him
    100 kbps per tv (+ overheads)... sometimes less... it's the lowest grade
    garbage I've ever seen but the area he lives in that's all that's
    available until Verizon comes online over there. I personally cannot
    understand how a provider can remain in business selling crap like that.



+ Reply to Thread
Page 2 of 2 FirstFirst 1 2