[G4] Broadband Speed Liar's Poker

[Erik Justus Paiewonsky]

Well for whatever it's worth - digest the following: I went to several 
of the test sites @ <http://home.cfl.rr.com/eaa/Bandwidth.htm> and with 
my ADSL connection here in Oslo, Norway ran several of the tests at 10 
randomly picked test sites worldwide. I ran each test 5 times or more 
in rapid succession, finishing with each test site test sequence within 
a minute before moving on.

Here are some interesting facts: These test sites measure a moment in 
time and give you no more accurate information than: "that at this 
particular few seconds in time" my connection speed was x seconds at y 
speed. I base my conclusion on the following: not only did none of the 
test sequences give me speeds that matched any of the other test site 
sequences, none of the test sites came up with the same connection 
speed twice in a row, from 10 seconds to the next 10 seconds.

Telenor, the Norwegian phone company that I subscribe to, avers that my 
account is a 704/105 account: (thus the "A" ("Asymmetric") in ADSL: 704 
kbps DL and 105 kbps UL). The results of the random tests (and let me 
remind you that I took each test at least 5 times in rapid succession 
before moving to the next test site, varied between 234 kbps to 2554 
kbps! More than a factor of ten difference!! Each site gave me a 
variance between 234 kbps-567 kbps at one site to 556 kbps-2554 kbps at 
another site. Averaging out all of my tests and keeping to the mean 
average, in other words, discounting the ridiculously low and the 
ridiculously high, I came to the conclusion that my average connection, 
or DL speed is between 500 kbps and 600 kbps, lower than the 704 
advertised, but certainly expectable. Given prior input from other 
'listers about overall traffic and that there will be slowdowns from 
servers on either end (and any along the way). Also don't forget that 
location in numbers of meters, (or yards, for those of you who are 
metrically challenged), from your local ADSL hub or switching center, 
and number of subscribers between you and the hub, also influence speed.

Bottom line: this is field where measurement is never going to be 
accurate. "How big is yours, take a look at mine!", just doesn't apply 
to connection speeds. Even within seconds, my connection speed 
registered huge differences performing the same test at the same time 
of day, with no other controllable variables on my part available to 
influence the results.

My advice: look at your average speeds over time and at different times 
of the day to see if you are being short-changed by your ISP. Other 
than that - Your Mileage Will Definitely Vary!!

On Sunday, Dec 14, 2003, at 14:13 Europe/Oslo, Power Macintosh G4 List 
wrote:

> Date: Sat, 13 Dec 2003 22:50:15 -0500
> Subject: [G4] Broadband Speed  Liar's Poker
> From: Snow White <jj4 at sympatico.ca>
> Message-Id: <A0833288-2DE8-11D8-B160-003065718F58 at sympatico.ca>
>
> I find many servers are slower than my 1.5 Mbps DSL.  I am in Canada
> and recommend Sympatico. Relliable and consistent.
>
> For speed tests go to
>
> http://home.cfl.rr.com/eaa/Bandwidth.htm
>
> There is about 80 or so places to check your speed all over the world.
>
>
> And I highly recommend DSL Reports for ISP experiences from other 
> users.
>
> www.dslreports.com
>
> Awesome site and you can test your Port Security as well.
>
> jj
>
Nosce te ipsum: "Know Thyself"
Inscribed at the temple of Apollo at Delphi

Erik Justus Paiewonsky
Oslo, NORWAY