RE: PBXs

["Steve Morgan" <steve@xxxxxxx>]

> >Businesses aren't replacing copper with fibre at office
> level, they're using it for the
> >backbone, particularly between buildings.

> Bollocks!! A very large number of companies are now running
> fibre to the
> switches, with some migrating it to desktop level.
> There's an article in Computer Weekly, Thurs 20 Dec, page 21 on it.
> That's the only one of many I've studied I have to hand at the mo.

Ooh, good start ;-)
I should, of course, have said Desktop, rather than office. My experience
is
similar to Keith's. I also work in an organisation that has some of the
most
complex computing requirements in the world. On our site there are probably
in the region of 2000 desktop machines and one of the largest data centres
in Europe. Countrywide, the network includes over 70,000 desktop PCs,
thousands of servers and a good number of mainframes. A very small number
of
these devices are connected by fibre. Fibre is used, as I mentioned, for
the
backbone, particularly between buildings within a campus. Desktop machines
are connected to the switches by CAT5. Outside of the data centres, most
run
at 10Mbps.

Compare, however, the quantity of fibre vs UTP in this kind of model. A
relatively small number of fibres connect comms equipment within a
building.
A very small number of longer fibres connect the buildings. Their use here
has nothing to do with bandwidth, but transmission limits. The buildings
themselves are flood-wired in CAT5. CAT5 rules supreme if you're talking
quantity in this kind of environment.

Question is, how can you apply this kind of model to a domestic
environment?
You can't, because corporate networks primarily involve clusters of large
numbers of low-bandwidth nodes, spread over a geographically large area.
It's a very different model to someone running a small number of
potentially
high-bandwidth nodes in a small area.

> >At home, though, what use is fibre? Copper is plenty fast
> enough for all but
> >the most esoteric of uses and is far easier to interface to.
> >
> For the moment that is the case, I haven't even attempted to claim
> otherwise. Once more I make the note that I am discussing the
> situation
> in 3 or more years time.

The thing is, if you install fibre now in anticipation of some future use,
how can you be sure it'll be compatible with the equipment you want to
install? I have two different reels of fibre-optic cable sitting in the
loft
that are no use to man-nor-beast, for that very reason.

> >How expensive would KAT5 be if it used fibre rather than
> CAT5? And how do you carry power
> >over fibre?

> Same answer as the above ... look down the line, not at the current
> scenario.

No, my point is quite a simple one. Fibre is used for the transmission of
light. Electrical signals (which we're still going to be using for a good
few years to come) can't be transmitted down a fibre without being
converted
into light (and back again). Fundamentally, that means that fibre requires
more complex interfacing that copper. For complexity, read expense.
Economies of scale will mean that for some types of devices (say, network
cards), that expense may become relatively small. But how cheap is the
fibre-optic version of an iButton going to be, or a fibre-optic CCTV
camera?

> >I think your assumption that fibre is the future of domestic
> networks is
> >flawed.

> I  clearly disagree with you, but my thoughts are based on
> accurate fact.

Hmm. Accurate facts relating to a different problem, I think.

> >People have been saying that fibre is all things to all men for
> >years and there's no evidence to suggest that they were correct.
> >
> There is a great deal of evidence to support the use [and indeed
> necessity] for fibre in corporate environments if you bother
> to look for
> it. I agree entirely that it has been overhyped and not suitable for
> every envirnoment/application .. but then, neither is cat5e.

Indeed, I have described a typical example of how corporates use the two
technologies.

> >I reckon that by the time we reach the point that UTP has
> had its day, something else
> >will be along to take its place and it probably won't be fibre.
> >
> Technologies like that take a long time to develop and reduce
> the cost
> to an affordable level. I haven't looked, but unless there is an
> alternative available right now [regardless of the cost] it
> is likely to
> be a long time before it's technologically ready and
> affordable for home
> use.

Wireless?

> >If I was undertaking a new install now and wanted to
> future-proof, I'd look
> >to install ducting to allow me to install new media as and when.
> >
> An excellent idea whatever cabling you settle on :-)

> >Putting fibre into the walls now would just be pouring money
> down the drain.
> >Just my opinion, like.
> >
> Yep, and the opinion of many others. But what of the cost of
> installing
> fibre in a few years? obviously for those without ducting ;-)

Therein lies the rub. It's the same as finding you've not installed enough
of your selected media to the appropriate places. It's the same as buying a
new Home Cinema system that requires yet more speakers around the room.

With something that is currently so limited in its application as fibre, I
don't see how you can decide how much to put where. It's too early to be
able to come to a sensible conclusion. Say you put in 8 fibres and 4 CAT5
cables behind your TV. What happens when the fibre-based stuff never turns
up but all 10 boxes under your TV have a UTP port? Media-changers really
are
relatively expensive.

It's an academic argument, obviously. Neither of us really have a clue how
we're going to connect everything together in 1,3,5,10 or 20 years time.
That's why, to me, the common sense approach is this: don't try to solve
the
problem now - make it easier to solve the problem when it arises.

I think you've caught the fibre-bug. Now, I'm off to see my Bank Manager
about starting a Ducting manufacturing company.

Regards,
Steve