🚀 go-pugleaf

I would guess that any cable company knows that:

a. the only medium term economic options for bi-directional broadband 
services to the home are cable modems, ADSL (over telephone cable), and 
LMDS/LMCS (in Canada)/MVDS/Europe (advanced wireless cable)

b. ADSL and LMDS are about 2 years away from mass roll-out covering all of 
the heavily populated parts of North America.  ADSL is sorting out standards 
and LMDS is still working on the technology (particularly the line of sight 
limitation).  This two year delay before competition is a major asset but the 
Cable companies (1) must get the technology right in light of their long-term 
future which must include going (completely?) digital (2) develop a uniform 
end-to-end technology/specs so that a real market can develop in equipment.  
This will reduce costs to them and also increase the chance of convincing 
subscribers to buy the modems thus substantially reducing their vast capital 
requirements.

c. that mass cable modem roll-out requires fibre to the node, nodes serving 
say 500 users and a bi-directional architecture.  There is a lot of 
investment, physical and human required to massively put into place fibre-to- 
the-node.  And Rogers, among others, will have some hard financing to 
arrange.

d. as the end-to-end cable modem specs will not be decided until late 1996 at 
the earliest (see below) I would think that many cable companies would  
concentrate, in the interim, in upgrading the human and physical 
infrastructure and lining up financing.
-----------------------------------------------
vanepp@wimsey.com wrote:
> 
> David Steinberg <amer2@hookup.net> writes:
> 
> >Would anyone care to comment on this question?  David Steinberg
> 
> >From:
> >        "Ciaran Clissmann" <ciaranc@net-cs.ucd.ie
> >I am working on data-over-CATV, and wonder how it is that CATV networks
> >are going to be able to deal with large amounts of upstream data, such
> >as that produced by videoconferencing, interactive multimedia, etc. The
> >small upstream capabilities of the cable modems I've seen (max claimed
> >speed 10Mbps), are not going to
> >be enough, it seems.
> 
> >What do you think ?
> 
> >Ciaran Clissmann
> >Teltec Ireland, University College Dublin
> 
>-----------------------------------------------------------------------------------------
> >Ciaran Clissmann, Teltec Ireland, Department of Computer Science,
> >Homepage construction site at htp://dallas.ucd.ie/~ciaranc
> 
>         The Rogers folks here in Vancouver (a year or so ago when we were doing
> a joint project to install a test cablemodem) stated their intent is to agregate
> around 2500 homes via the cablemodems back to neighborhood hub. The hub would
> then tie in to a FDDI ring (I expect these days the hub will tie in to an ATM
> network) to handle the aggregate data. There are of course still potential
> performance problems if everyone piles on at the same time, but at least they
> aren't planning on running the complete plant on cablemodems. I have also
> heard rumors that the cablemodem rollout will be delayed a year, so perhaps
> things weren't as easy as expected.
> 
> Peter Van Epp -------------------------------------------
re standards
wbmaster@rogerswave.ca

1. I understand that a consultant company is carrying out a study likely
> to result in cable modem and related standards throughout Canada and the > USA.
>
> * Not really.  A.D. Little is providing assistance along with a group of > vendors & MSO's.  Cablelabs and DAVIC are also helping to a small
degree.

> The real standards decisions will be made by the IEEE 802.14 committee.

> They should make some major decisions towards the end of this year.  Note a standard will be a global standard and not just one for the us
and canada an iso.
>
>   Do you know how it is progressing and whether/when we will have
> (end-to-end?) standards for cable modems and all the supporting systems?
>
> *The IEEE 802.14 group should make some major decisions towards the end of this year.  There will be an end to end standard.
>
>  Would I be correct to assume that the proposed standards would centre
> on 30 MBS cable modems leaving 10MBS as a never produced standard?
>
> *No, that would be a bad assumption.  The committee is not focused on a standard that will provide the highest possible data rates.  They are
instead concerned about the most technically sound solution.  There are
legitimate problems with the return path and with bandwidth allocation.
These issues are more important than the data rates.  I would guess that
they would go with an assymetrical, atm protocol and qam/qpsk
modulation.
> The still in the lab gi/fore atm modem does get 27Mbps downstream, but who knows if it works.
>
> 2. On the otherhand what is happening re.:
>
> a. defining the technology, standards and inter-operability of set top
> boxes?
>
> *The IEEE802.14 is the committee that is responsible for the itv set top box.  It is on the backburner.  After the cable modem standards are
resolved, then they will return to the set top box.  DAVIC is
responsible for the ADSL solution and they are definitely making more
progress than the cable industry.  Interactive tv is at least 5 years
away if it ever shows up.
>
> >b. what is happening re digitizing the cable tv systems?
>
> *not much
>RE ATM STANDARD
>I do think that atm will be the standard for transport protocol because with 
atm you can dynamically allocate bandwidth.  You cannot do that with 
ethernet.  Bandwidth will not be a near future problem for service providers 
but will be in 3 years or so.  There are several ways that bandwidth can be 
hogged (a perfect example is running .exe files out of netscape) and one user 
may be able to slow the whole bunch down.  

There have been trials using an atm platform.  In fact, Cox released a press 
release today that they will be conducting an atm trial.  More and more 
manufacturers are going towards atm or a multi-protocol modem that will work 
with an atm based network.  

The opposition to atm is primarily aimed at the cost and "newness" of the 
technology.  There are few that can point out actual technical flaws with it. 
 I noticed that when you mentioned it, you only stated that it had not been 
used much in the field.  I really don't think that field experience will have 
an effect on the standard.  Zenith has more modems in the field than anyone 
and there is not a chance in hell that BPSK will be used as the modulation 
scheme.  The real issue is whether or not the elite engineers at the top who 
will be making the decisions will have faith in such a young
technology.