On Aug 21, 9:20 am, jcon wrote:
On Aug 21, 3:21 am, Poutnik wrote:

The correction was not done initially,
but later these smart engineers realized there is need to introduce it.
It does not matter in what part on system it is done.

Not quite. Scientists (of course) knew the correction would be
needed,

The correction might be needed if acquiring GPS almanac information
from only 3 satellites. shrug

but some engineers doubted it, so the first satellite had the ability
to *toggle* the (major) correction on and off.

Because the engineers saw that by acquiring data from at least 4
independent satellites, there is no need to correct anything even if
the anomaly existed. shrug

It was initially
operated
for 20 days without it, the predicted shift was observed, and then it
was turned on and left on:

That was testing GR predictions. It had nothing to do with the GPS
other than tested under a proto-GPS satellite. shrug

What GR predicts is also predicted by other hypotheses. One example
is to allow the speed of light to vary according to the following to
the first order.

C(r) = c0 (1 – G M / c^2 / r)

At higher altitude, the speed of light is higher. Thus, any
mechanisms that scores time goes faster by almost exactly what GR
predicts. shrug

However, the killer is the SR part --- namely this 7usec stuff. This
amount should apply equally to both the satellites and the receiver.
In actual applications, the necessary correction is only done one
way. Thus, GPS definitively proves GR wrong. shrug

http://www.phys.lsu.edu/mog/mog9/node9.html

Ashby makes a rather unfortunate statement in the paper.
he says the GPS system "can no longer be used to test general
relativity".

Given him the benefit of the doubt, he actually understood that GPS
does not prove the validity of GR since he said it in the first
paragraph that by acquiring almanac data from 4 satellites, the
critical time information as measure in satellite time can be solved
every time. shrug

Because
the clocks are now synchronized with ground clocks fairly frequently,

Synchronization is basically done by resetting a time-keeping
counter. It is done so for obvious reasons in which Professor Ashby
has mentioned later in the article. shrug

This is often misinterpreted to mean that the
relativistic corrections are not important.

Are you kidding? Einstein Dngleberries still believe in the myth that
relativistic effect is needed in GPS. Have you not read the posts
from these Einstein Dingleberries? shrug

They are.

No, they are not. Professor Ashby disagrees with you. shrug

The statement simply means
that because the system if frequently synchronized, it cannot improve
on the tests which were initially done.

The relativistic effect only accounts for 450 parts per trillion of
error. There are other parameters that account far more than that.
Since synchronization of time, regardless how oscillation frequency
varies, is done through software algorithm such as IEEE1588 as an
example, there is really no need to make sure each oscillator achieves
a 450 parts-in-a-trillion of accuracy. Do you know how expensive to
achieve and test for that are? shrug

http://en.wikipedia.org/wiki/Precision_Time_Protocol