lcliffe@gmail.com hath wroth:
>They are accessing the same access point so the model and revision of
>the AP doesn't make any difference. It is not a bit/byte error because
>that wouldn't be a third now would it? It'd be more like an 1/8 (not
>exactly I know).
The reason I ask for numbers and methodology is that there are a
substantial number of errors that can be attributed to arithmetic and
proceedural errors. My basic proceedure is to isolate the potential
culprits by elimination or replacement. That's a bit difficult to do
without some clue as to what you're working with, what your doing, and
what you're getting. For example, setting your unspecified access
point to a fixed speed instead of automatic has a big effect on
performance in the presence of interference. Turning OFF the 802.11b
compatibility feature is another. Disarming the Super-G and
Afterburner functions seems to also change performance somewhat.
>And as for software to test, it doesn't matter. It's the same software
>on both machines.
I'll be blunt...(when am I not blunt?). It's difficult enough to
deduce the cause of a problem without additional handicaps. Some
benchmarking methodologies return useless numbers. For example, using
Firefox to download and save a file from the internet results in
numbers that are affected by the time delay between when the download
starts, and when the user elects to hit the "save" button. Other
benchmarks are severely affected by local cacheing. I recommend
IPerf:
http://www.noc.ucf.edu/Tools/Iperf/
>Does anybody know of a reason why two identically configured machines
>would have different wireless download on a consistant basis from the
>AP regardless of whether they are both accessing the AP at the same
>time or not.
Sure. I can think of lots of reasons. One of my friends calls me up
and declared his new wireless laptop to be slow dog. Two hours later,
he calls to inform me that it magically has fixed itself and is now
running at normal speed. The Windoze background downloader was
merrily filling his DSL pipe with updates without the slightest
indication on the screen. Needless to say, it was slow during the
updates.
>There is no difference in the machines despite what your
>hunch may be,
You seem to be making a substantial number of assumptions. Have you
ever considered testing your assumptions? For example, have you
considered swapping "identical" Intel 3945abg cards? Have you
consider dragging the slow laptop over to another access point and
seeing if the problem persists? If it doesn't, then there's some
incompatibility or settings problem between the access point and
laptop.
>I've spent two days getting every driver, firmware, MS
>patch to be equal on both machines.
Just answer one question (for my amusement). Are you using Intel's
Proset utilities to manage the wireless or are you using whatever Dell
supplied?
>The problem is limited only to the
>wireless connection and is completely repeatable.
Yep. I've noticed that if I make the same mistake over and over, it
usually results in the same problem, over and over. Check your
assumptions.
>Sorry if I sound sore about the whole thing, I'm just a bit frustrated
>with it.
That would be an undestatement. I presume you've also called Dell
support to vent your frustrations. They might have a BIOS update for
the 9400 that will magically fix things. That's what happened to me
with an older Dell 5150 laptop and the 2200bg card. It was doing
weird things. Replacing the card and screwing around with Windoze did
nothing. I had the latest BIOS installed, but the product had just
been released and I thought it might be worth a call. Apparently,
they were working on the exact problem and sent me a beta BIOS image.
It worked.
>I know that generally we have to start at the basics, but
>I've been doing this for years, I've covered the basics many times
>over.
I kinda like reading Sherlock Holmes mysteries. One of stories had a
line something like "When you have eliminated the impossible, whatever
remains, however improbable, must be the truth." If you can't point
directly at the problem, take the time to eliminate what is NOT the
problem. For example, replace the access point and see if there's a
differnce. Swap the wireless cards. Take the laptop to a different
location to reduce interference problems. Try different drivers
(Proset versus WZC).
>The information I've given is a good summery, actual numbers
>won't help because they don't impact the problem at all.
That which is most obviously correct, beyond all need of checking, is
usually the problem. Check your assumptions.
--
Jeff Liebermann
jeffl@comix.santa-cruz.ca.us
150 Felker St #D
http://www.LearnByDestroying.com
Santa Cruz CA 95060
http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
Weird Wireless Speeds 
Linear Mode
