The medieval idiom "Third time's a charm" was part of the writer's motivation
for a third try to find hardware and connectivity profiles that could be recommended to new
and current OH internet users to give optimal internet access speeds. This third try, too, failed.
The U.S. baseball idiom "Strike Three, You're Out" is
appropriate. No more attempts to create such profiles will be undertaken.
First Time, Strike One
When the OH IT Director reported that the major cabling upgrades where completed, requests for the same
speed test were sent to the 60. The number responding with the same test totaled 49: the first was on 9/1/19,
the last on 10/30/19. These gave the "After" speed tests. Participants were asked to do the After tests the
day of the week they did their first at approximately the same time.
Detailed analysis of the Before and After tests was undertaken with Excel. Multivariate models of factors
that might explain individual users' differences in test speeds were created. Regression tests with Excel gave
NO significant results at conventional levels (.001, .01, .05). It was found that OH residents connecting micros
to their Ruckus modems had faster speeds than those using only wireless. This was in keeping with what has
been known for many years in the network industry.
These First Time results were made available to the 49 OH residents doing the Before and After tests. They
are now available at the following Web sites and can be opened using a browser:
OHInternetUpgrade11.pdf
628KB (11 paper pages) [1]
exhibits11.pdf
1,346KB (11 paper pages) [2]
Opening either gives a prompt for a username and a password. Both are equal to "flowersunfl" typed without
quotes in the two fields. After they are typed, click on the "OK" button under the password field. If the
chosen image does not open, please let me know. With an open image, place your cursor arrow on the PDF,
hold down the Windows Ctrl or Apple Command key and press plus ("+") to make the image larger. Press X or
Back to exit.
Do not share this Username/Password with anyone. It and the included information and links must be
treated as we have agreed to treat our Residents Living at Oak Hammock listing available from the OH Lead
Receptionist. It states, "This information for Oak Hammock use ONLY. Not to be distributed outside of Oak
Hammock at the University of Florida."
The Speedof.Me test was chosen, in part, because it returned 4 tests to the user: Download Maximum,
Download Average, Upload Maximum, and Upload Average on up to 11 different file sizes. The speed for the
last file taking 8 seconds is reported as the "Average". Ten reasons why it was used are listed on Page 2 of [1].
The mean speeds for our 49 participants were very significantly higher for all 4 types of tests (rounded to
millions of bits per second next):
Details on these mean measures and their standard deviations are in [1], Section 6, Tables 1 & 2 with all
participants' results in [2]. The standard deviations show MUCH more variation among the 49 OH participants
in the After sample.
Second Time, Strike Two
The large variation among OH participants led to the idea users' configurations of keyboards to Ruckus
routers could show what gave higher speeds and what gave no or lower speeds. Following some analysis of
Before and After speeds, it was found that percentage CHANGE of After speed versus Before for each
participant permitted a logical classification of individual users' download averages.
A. Nine of the 49 participants had no or negative changes.
Following review of the classification structure with 4 OH residents Bill Castine, Tom George, John Paul and
Bill Zeger a tentative plan was created to form a Hammock InterNet Team (HINT). It would interview
participants in
A. & B. above to create keyboard-to-Ruckus schematics in their residences. Then, summarize them in a form
that would permit the HINT to formulate the desired profiles. HINT would include the 4 individuals noted in
the prior sentence, plus Bob Stott, OH IT Director, and Roy Padgett, IT Networking GRU and the author.
Teams of 2 from HINT were to be arranged for interviews in participants' residences with at least one
familiar with Windows or Apple systems depending on a particular resident. The estimated time per interview
was expected to be 30 to 45 minutes, depending, again, on a particular resident's system complexity. This
plan, was reviewed and approved by the OH CEO and CFO (as was the first).
Then Covid 19 arrived! The OH "closing" rules, still very successful from the viewpoint of OH residents'
health, did not permit the interviewing plan to take place. Strike two!
Third Time, Strike Three
1. Create a plan to get residents' system schematics with no direct person-to-person contact.
With 22 months of effort then invested in finding the now-elusive optimal Windows and Apple profiles,
giving up, Alternative 3, was not yet acceptable. We (and the rest of the world!) entered and
ended the summer months with a second CV-19 phase presumably on the horizon and no vaccines were in place.
Alternative 2, waiting for the end, has
no horizon for all practical purposes. Alternative 1 became the default option at month 30 with OH becoming
less closed.
A keyboard-to-Ruckus schematic for the author's setup at 2428 OH was reviewed with OH members of what
was to be HINT (as noted above). They gave fixes that were included in the authors' schematic. It can be
opened with the following URL in a browser using the Username/Password flowersunfl:
2Elnicki.jpg
116KB (1 page) [3]
Four OH members of HEAT were asked to create similar schematics for their resident setup and drawn
by hand on paper. While
there are many graphics software packages, both for free and for a fee, acquiring, learning to use one, and using it
to draw one schematic would not be a feasible request. An email was sent to the 9 OH participants in the 49
with the fastest absolute After speed tests. It contained a request for hand-drawn schematics of the keyboard-
to-Ruckus systems in their residences. The request included a "legend" listing the kinds of devices that could
be included in the schematics:
legend1.docx
14KB (1 page) [4]
Five responded: 3 are members of the OH HEAT group, 1 by a HINT member, and 1 OH resident that
permitted the author to create a schematic for him. This resulted in 10 schematics including 5 from the HINT
group.
This led to the conclusion OH residents not in HEAT or (the tentative) HINT had no interest in supporting this
effort. "Strike Three, You're Out" was concluded (as stated in the first paragraph above). The 10 schematics
have been "summarized" in an Excel spreadsheet. It is in a PDF format and can be seen at
schematic10.pdf
90KB (11 pages) [5]
They illustrate the very heterogeneous nature of keyboard-to-Ruckus setups among OH residents. This is
consistent with the much larger variation found in the After tests and compared to the Before tests. Each can
be viewed individually at the following site.
scans.htm
116KB (1 page) [6]
Most cannot be read in the sizes given by [6]. They can be read by enlarging each. To do so, click on one of the 10
names to open a schematic. Place your micro's cursor on the opened schematic, hold down the Windows Cntl
or Apple Command key and press the Plus ("+") sign repeatedly. The image will get large enough to easily read
the components. Click on your micro's Back or Exit command to return to the display of 10 schematics in [6].
The first attempt began before the start of the OH internet connectivity to upgrade all residents' internet
access. The first of 60 participating residents responded with the requested speed test on 4/24/18, the last on
7/20/18. These gave the "Before" speed tests using Speedof.Me.
Direction Download Upload Measure Maximum Average Maximum Average Before Means 47 37 66 58 After Means 152 133 180 169 Ratios 3.3 3.6 4.2 4.4
B. Fourteen with speed changes greater than five-fold had Before speeds under 50.
After the two failures, the author debated about the reasonable alternatives available.
2. Wait for the CV-19 to be gone or fully controlled.
3. Give up.
/* Send Email to Dick
Last Updated 11/20/20
C:\HEAT\ohend.htm