On 26 Apr 2007 18:30:06 GMT, thanatoid
<waiting@the.exit.invalid> wrote:


You're at /least/ a little confused.

http://www.google.com/search?hl=en&q=define%3Achange

On Thu, 26 Apr 2007 12:09:23 -0700, UCLAN <nomail@thanks.org> put
finger to keyboard and composed:

I use Agent. If you fiddle with the subject header, then Agent will
start another thread.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

~ Is it possible to obtain a 56 Mbps connection using 1,000 dial-modems
~ with a 1,000 different telephone ports and numbers?
~
~ Each line gives a max of 56 Kbps, so if 1,000 are used, could this
~ give a max of 56 Mbps?

Well, let's gloss over the issue of whether a modem link really gives
you as much as 56kbps (it really *can*, in the digital->analog direction,
in rare instances), or whether it gives you more due to compression
(it can, if the data has low entropy, but that in that regard a modem
link is no different from any other data link), and rephrase the question:

Is it possible to obtain an x * n bps connection, using x links, each of
n bps?

The answer is, maybe, depending on x, n, and what you mean by "bps".

Some options for aggregating multiple links are:

inverse multiplexing ("bonding")

multilink PPP

equal-cost multipath routing

The load-spreading characteristics of the different link aggregation methods
vary. Inverse muxing spreads the load most smoothly; equal-cost IP routing
the least. In order to take advantage of all member links, your offered load
may have to provide multiple outstanding PDUs, and even (with some equal-cost
IP routing algorithms) multiple concurrent flows.

Does it make sense to bundle 1000 dialup modem links together?, to get a
"56 Mbps" point to point link? No. 1000 POTS lines would cost you some
US$40,000/month on each side, not to mention the power consumption, not to
mention the inherent nastiness of managing a rack of 1000 modems. (I
actually used to manage such a system, so I know what I'm talking about.)
This would be a poor use of $40,000/month.

Regards,

Aaron

On Apr 26, 3:58 pm, Aaron Leonard <A...@Cisco.COM> wrote:

Thanks for clearing this up.

I was just asking whether this was hypothetically-possible. Even if I
had Bill-Gates's $$billions, I still wouldn't actually do this. I am
just interested in whats possible in theory.

I use a Cable connection for the net. I used to use dial-up around 6
years ago. Ever since, I quit dial-up and switched to cable because it
is much faster.

However, I still like the sounds associated with the dial-up modem. I
briefly hear this sound if I pick a up telephone in the same line as
the dial-up modem is using and can hear similar sounds as the modem
connects to the net [a few second after I pick up the phone and hear
this sound, the PC's modem disconnects from the internet].

Dial up modems operate on the same frequency as regular voice
communications on the phone line - between 300 and 3000 Hz. Does this
mean that sounds on the line used by dial-up modem have a maximum
pitch of 3 KHz?

"S" <s@example.net> wrote in
news:4630fe81$0$97257$892e7fe2@authen.yellow.readf reenews.net:

Neither precludes the other.

That being said, can we change the name of the thread to "anal-
retentive geeks with nothing to do talk about totally irrelevant
things"... Oh wait... That would pretty much apply to 95% of
Usenet text posts, wouldn't it...

I am old and my memory is fading daily and at *this* moment I
don't feel that this is worth going back and searching through
all the posts in the thread which I curse myself for having
joined (although it HAS given THIS anal-retentive geek with
nothing to do *something* to do).

I do however basically agree with the statement "compression is
how modems achieve their rated speeds". This statement may be
wrong. I really don't care. Nonetheless, if you have as little
to do as I, give me a non-tech link which explains in plain
English how/why this is NOT the case.

"which is wrong" is not good enough, sorry. Even with two spaces
before it.

See start of above paragraph, but I do not believe I ever said
that. AFAIAC, "LINE" does not have a "speed". It is a piece of
wire.

Like I said, if ANY of this really mattered to me, and if I was
a 21st Century Speed Freak I *would* have BB. I CAN afford it.
It's just something I do NOT need. Sigh.

As I just stated, I don't think so. Does it REALLY matter?
If it DOES to you, quote me directly.

I admit I am wrong when someone proves it to me logically or
factually in "regular" English. At least one or two of those 3
elements are missing from most Usenet posts, including this one
of yours I am replying to. You could at least have used quotes
to back up what you claim I said or believe.

And, again, does it REALLY matter anyway?

As I gently suggested in a previous post, "troll methods" are
just a normal part of human/semi-human communication which some
people with 12 MB hard drives up their asses seem to arbitrarily
have occasional problems with.

(Speaking of trolls, the REAL troll in this mess, the OP, was
EXTREMELY successful in baiting us to drag this on for this
long. S/he must be peeing his/her pants.)


--
Disagreements and the usual insults expected and welcomed.

On 27 Apr 2007 01:44:40 GMT, thanatoid
<waiting@the.exit.invalid> wrote:



Do you not even have the slighest fathom of why a thread
subject line is to stay UNCHANGED?




Wrong. "Troll methods" are employed by those who have no
useful purpose, who only aim to cause waste or elicit
emotion from others who had a useful goal.



Then why did you take the bait?

One the very first post by the troll, a trend is set. We
could see instantaneously that there was no purpose. Of
course if one wrote an elaborate enough proxy, all these
modems could be teamed to achive a higher bitrate. Even so,
it would be an incredible waste since broadband exists and
could achieve so much higher with far fewer slave systems.

To concentrate on an unrealistic scenario, the OP began with
an obvious attempt at unfruitful discourse.

kony <spam@spam.com> wrote in
news:dm1333t7s1qida76p3fqlsd51mvb3jees4@4ax.com:

Do you not even have the slightest amount of a sense of humor?
As to changing thread subject names, I believe another
respondent put it rather succinctly... Check again if you
haven't seen it.

You REALLY should try to get a life.

I have nothing better to do. Yes, *I* have NO life, which
doesn't mean everyone should be like me.

And so the OP keeps on peeing his/her pants...

Next?


--
Disagreements and the usual insults expected and welcomed.

S wrote:

I've been using T-Bird for over two years and never had that occur. If
the originator of the new "Troll Thread 1" message uses "WRITE" instead
of "REPLY", a new thread will result.

Franc Zabkar wrote:

Really? Even when "REPLYING" to a threaded post? Sounds like a problem
in Agent. What happens if you enter a NEW message with the exact same
SUBJECT as a current thread? Should start a new thread.

kony wrote:
I firmly disagree. The subject should describe the material. When
the subject changes there is no real reason to maintain access to
older posts, although the reference system will preserve all that.

--
<http://www.cs.auckland.ac.nz/~pgut001/pubs/vista_cost.txt>
<http://www.securityfocus.com/columnists/423>
<http://www.aaxnet.com/editor/edit043.html>
<http://kadaitcha.cx/vista/dogsbreakfast/index.html>
cbfalconer at maineline dot net



--
Posted via a free Usenet account from http://www.teranews.com

On Fri, 27 Apr 2007 09:18:46 -0400, CBFalconer
<cbfalconer@yahoo.com> wrote:


If the content of the post was deviating from the original
subject, yes. If it is just a typical thread reply where
the poster replied as everyone else had, but felt they
should edit the subject line to inject some personal opinion
that could have been in the body of the post, no.

On Thu, 26 Apr 2007 23:25:22 -0700, UCLAN <nomail@thanks.org> put
finger to keyboard and composed:

Here's an old post of mine:

Aus PC User modem shootout (was Zoom or US robotics modems)
http://groups.google.com/group/comp....6dd3634?hl=en&

IIRC, this post showed up in my newsfeed as a new thread. However
Google Groups lumps it with the original thread with the following
notification:

"Discussion subject changed to 'Aus PC User modem shootout (was Zoom
or US robotics modems)' by Franc Zabkar"

I'm not sure what happens, but I'd rather not annoy the group by
experimenting. However, John Navas' periodic reposts of his FAQ end up
in the same thread. In fact I have two copies of the xDSL FAQ (at
comp.dcom.xdsl) dated 21/06/04 and 19/04/07.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

On 24 Apr 2007, in the Usenet newsgroup comp.dcom.modems, in article
<1177464315.951399.266160@t39g2000prd.googlegroups .com>, Radium wrote:


OK - think of an IP connection as you having a conversation with some
other person. As long as there is one IP address at each end of
the conversation (bonding), you will see an increase. If there are
instead, one hundred addresses on your end, your peer in this
conversation is going to be EXTREMELY confused. "I'm trying to
talk to you, and all these other people keep butting in to the
conversation." So that ain't gonna work. Have you got a web
page you're trying to load with 100 different images built in? As
long as there are no inter-dependencies (cookies and such) then you
may be able to download them over separate links, but your operating
system is going to go nuts trying to figure out who (which application)
gets what, and I don't know of any operating system and application
such as a browser or proxy server that can handle this.

Yes, each connection can carry independent data - the problem is trying
to consolidate the resulting mess into a coherent picture. If you are
old enough to remember the Applo 11 landing on the moon (or going way
back - the coronation of Queen Elizabeth in 1952) there was very limited
world wide television coverage - the 1952 event took _all_ of the under
sea cable telephone circuits to deliver a crappy picture. We don't do
that any more, because there are better ways (satellite relay). No one
is wasting time trying to do it the old fashioned way.

Old guy

Radium wrote:
No, you cant ever even get a 56 connection with a single modem on a
single computer.

--
http://www.bootdisk.com/

"Radium" <glucegen1@gmail.com> wrote in message
news:1177631348.493180.165980@b40g2000prd.googlegr oups.com...

yes. the American system may be 300 to 3400 hz, i'm not sure.

C = 0.332 * B * SNR
C = 0.332 * 3000 * 30
C = 29,880 bits/second

in that approximation, bandwidth and signal to noise ratio are proportional
to capacity.

"T" <t@example.net> wrote:
It is intended to be "at least" 400-2800Hz. That is the
*minimum* specification for a "voice grade" dilaup line. (There
are several different "voice grade" specifications, including
400-2800Hz (dialup), 300-3000Hz (private line), and 300-3400Hz
(PCM link), for different types of equipment.)

However, 1) most lines are significantly better than minimum
specifications, and 2) if there is no way to provide at least a
minimum specified line the telco can legally opt to provide a
below minimum specified line that will actually work well enough
to provide useable voice calls.

Note also that the minimum Signal to Noise Ratio is 24 dB, but
typically would be much better than that.

Shannon's formula is

Capacity = Bandwidth * ( 1 + S/N )

where Capacity is bits per second
Bandwidth is channel bandwidth in Hertz
S is signal power
N is Gaussian noise power (expressed in the same terms as S)

Note that S/N is *not* expressed in terms of decibels.

But if the SNR is significantly greater than 1, the above formula
is approximately correct where B is Hz bandwidth and SNR is expressed
in dB.

And... much more interesting numbers can be plugged in...

A 400-2800Hz channel with a 24 dB SNR,

C = 0.332 * 2400 * 24 = 19,123

Keep that figure in mind for the discussion of v.32bis below.

Bandwidth of a typical telephone connection varies from a
minimum of 400-2800Hz, up to something approaching 0-4000Hz, or
even greater under some circumstances. Signal to Noise ratios
are specified at 24 dB minimum, but might be as high as 65 dB.

Connections that include any type of digital carrier
transmission system, including digital switching systems, will
have less than 0-4000Hz bandwidth, and an SNR no higher than 37
dB; but on a directly connected wire loop using a typical
mechanical switching system, the bandwidth will be greater (T1
lines require 750KHz, for example.)

Using the approximation of Shannon's formula and the minimum
specification for a PCM link, and assuming the cables connecting
each end add as much noise as the PCM link, we get

C = 0.332 * 3100 * 34 = 34,992

as the absolute best one could ever get through a digital
switched PSTN connection. Which explains why v.34++ only goes
to 33.6Kbps.

For connections that only include some form of analog carrier
transmission system, the bandwidth will almost certainly be
limited to 200-3500Hz at the most, but might have SNR values
greater than 45 dB too.

For connections that include some form of standard 64Kbps PCM
digital carrier, the bandwidth is limited to a maximum of
80-3750Hz, and the SNR will never be more than 37 dB. (For
other digital systems such as any of those using 32 Kbps, the
ranges are lower.)

Hence the actual bandwidth and SNR of any given telephone
connection can vary greatly depending on how the connection is
routed. Modem protocols such as V.34 were designed to take that
into account and provide the best data rates possible for
whatever conditions actually exist on the line. Prior to that
all protocols were designed only to take advantage of a
minimally specified telephone line (hence a v.32bis modem can
obtain a 14.4Kbps rate over a mimimally specified telephone
connection).

A few requirements for various data rates and protocols are,

Rate Protocol Bandwith Frequency_Range
14.4Kbps v.32bis 2400Hz 600-3000Hz
24.0 v.34 2800 467-3267
26.0 v.34 3000 375-3375
28.8 v.34 3200 320-3520
31.2 v.34 3200 359-3559
33.6 v.34 3429 244-3674

V.90 protocols require 3600Hz of bandwidth from
150 to 3750Hz within the following limits,

Less than 10 dB rolloff below 300Hz
Less than 24 dB rolloff above 3450Hz
Less than 43 dB attenuation end to end
No load coils, bridgetaps or wire gauge changes
Only 1 digital to analog conversion (CODEC)

What signficance is there to all of that??? Well, it means that
a typical v.90/v.34 modem makes a *very* good piece of test
equipment to determine the quality of a telephone line! If a
14.4Kbps connection can be established, the line clearly meets
minimum specifications. If it can accomplish any of the higher
v.34 bit rates, it is clearly a very good line. A connection
that allows negociation of any v.90 bit rate is simply an
astounding feat!

And, after all of those numbers for bandwidth... a word of
caution: Shannon was talking about 6 dB bandwidth points, but
the 400-2800Hz minimum spec for a telephone line can have as
much as 14 dB rolloff at either 400 or 2800Hz compared to
1000Hz. So all of the above bandwidth figures are relative.

--
Floyd L. Davidson <http://www.apaflo.com/floyd_davidson>
Ukpeagvik (Barrow, Alaska) floyd@apaflo.com

"Fred Hess" <tin-canned-courtesan@displeasing-fish-hole.net> wrote in
message
newsA5E1AC9.924641F3@alt.recovery.thats.not.milk .thats.pus.you.bitch...

feel free to do all the trolling you want, one day you'll wake up and
realize you are a loser and a nobody.