In a previous thread ("Einstein's Relativity contains a HUGE Loophole. Its
Implications Can't Be Ignored") a common assumption is made that clock
synchronisation is essential to setting up and defining special relativity.

Then an argument is presented critiquing the relevant clock synchronisation
(typically, Einstein's) as in some ways deficient or arbitrary or impossible to
verify experimentally. This then "leads" to the conclusion that special
relativity is circular at best and self-contradictory at worst.

But this is an incorrect, although widespread, assumption. Special relativity
is a theory which is based on a geometry (whether this is a sensible thing to
do is another topic) and therefore it does not need coordinate axes, or clock
sync, for its definition. It's the same with Euclidean geometry: it existed
just fine before Rene Descartes.

So I'll describe below the relevant experimental setup that necessarily implies
special relativity holds (to the experiment's error bars) WITHOUT EVER
referring to, or relying on, synchronising clocks. It only assumes a bunch of
rulers (identical in construction) and a bunch of clocks (identical in
construction). The rulers have uniform gauges engraved on them identically
(when positioned next to each other), the clocks tick identically (when
positioned next to each other).

We perform the experiments described below far away from any external forces
(including gravity) so objects when left alone travel uniformly in straight
lines.

We first establish that the geometry of space is Euclidean. This means we need
to verify the Pythagorean theorem. Which in turn means we need to
experimentally set up the model for "right triangle".

This is easily arranged: take two rulers of the same length (experimentally
verifying beforehand that they are congruent) and arrange them in line, end to
end. Let's say the first ruler (+++) is AB and the second (===) is BC:

++++++++++==========
A B C

Now we pick a point D not on that line that's equidistant from A and C. Again,
"equidistant" can be verified with our rulers. Then we call the triangle ABD
"right triangle" and we claim the following (Pythagorean theorem):

AB^2 + BD^2 = AD^2 (*)

(same thing could be done using the triangle CBD).

If the result (*) holds for any initial configuration of the points A, B, D set
up according to the above procedure, then we'll have established the spatial
geometry as Euclidean.

Now comes the less obvious part: establishing the Minkowski geometry when time
is added as an extra degree of freedom.

This means verifying a "Lorentz" version of the Pythagorean theorem which
involves two cases: the hypoteneuse being measured by a clock vs. the
hypoteneuse being measured by a ruler.

Here is our experimental setup:

1. a light source/detector and a clock (all three co-located). Let's call this
clock "clock 1".

2. a ruler with a mirror mounted on it (pointing in the direction along the
ruler), say at the "0" mark of its gauge engraving. We also put one of our
clocks at that mirror ("clock 2").

The clocks are not synchronised or offset in any particular way.

We perform two types of experiments.

EXPERIMENT 1 ("the hypoteneuse measured by a clock")

Put the light source somewhere on the ruler and set it in a uniform motion
along it. Do it so that at some instant the light source and the mirror
coincide. Take their clock readings at that instant. Let's say those readings
are C1 and C2. Meaning: clock 1 reads C1, clock 2 reads C2.

Some time after that, let the moving light source emit a light pulse toward the
mirror. Suppose the light source clock (clock 1) reads C1' at the instant of
the light pulse emission.

The light travels along the rules and gets reflected by the ruler's mirror at
its time C2' (clock 2).

The light pulse finally gets back to the light source when its clock reads C1''.

We notice that the numbers so collected always obey:

(C1' - C1)(C1'' - C1) = (C2' - C2)^2 (**)

EXPERIMENT 2 ("the hypoteneuse is a ruler")

For this part of the experiment we need to account for the arbitrariness of our
choice of the rulers' gauge and the clock ticks. We first find out what the
two-way speed of light is, measured only once, in an arbitrarily fixed system
(as long as it moves uniformly, as we stipulate throughout).

A two-way measurement like this again does not require any clock
synchronisation and we do not even assume anything special about this: no
"universality" or "constancy" of this number, no "second postulate", none of
this. Just measure the two-way speed of light, just once. Call the resulting
number "V" (in deference to Einstein's 1905 paper).

Now we perform a similar experiment. We begin as in Experiment 1 by setting the
light source in motion along the ruler. This time we don't bother recording the
two clocks' readings when the light source and the mirror coincide. We just let
the process continue.

And again, we emit the light pulse at C1' toward the mirror. Now at some
arbitrarily chosen instant but BEFORE the reflected light pulse hits the light
source/detector again, we write down two things concerning clock 1:

- its reading, call it C1*,
- the ruler's engraved distance marking at that clock position at that instant,
call it D*.

Then we write down the clock 1 reading when the light pulse hits the detector (C1''),
as before.

We notice that the numbers so collected always obey:

(C1' - C1*)(C1'' - C1*) = -(D*)^2/V^2 (***)

That's it. If the numbers collected in the above experiments satisfy (*), (**),
and (***), then space and time must necessarily obey the entirety of special
relativity (time dilation, twin paradox, etc. etc. etc.). Of course all that up
to the experiment's error bars.

Of course this method of defining special relativity suffers from exactly the
same shortcomings as the
pre-Descartes geometry. But THE POINT here is that clock synchronisation is
never used or referred to here.
In other words, clock sync is NOT any essential aspect of special relativity,
let alone a defining one (as it's frequently claimed). Such synchronisation, as
well as the notion of simultaneity that follows, is thus only a convenience for
the theory. Just like the choice of coordinate axes is only a convenience for
Euclidean geometry.

--
Jan

On 02/11/2024 01:08 PM, JanPB wrote:
> In a previous thread ("Einstein's Relativity contains a HUGE Loophole.
> Its Implications Can't Be Ignored") a common assumption is made that
> clock synchronisation is essential to setting up and defining special
> relativity.
>
> Then an argument is presented critiquing the relevant clock
> synchronisation (typically, Einstein's) as in some ways deficient or
> arbitrary or impossible to verify experimentally. This then "leads" to
> the conclusion that special relativity is circular at best and
> self-contradictory at worst.
>
> But this is an incorrect, although widespread, assumption. Special
> relativity is a theory which is based on a geometry (whether this is a
> sensible thing to do is another topic) and therefore it does not need
> coordinate axes, or clock sync, for its definition. It's the same with
> Euclidean geometry: it existed just fine before Rene Descartes.
>
> So I'll describe below the relevant experimental setup that necessarily
> implies special relativity holds (to the experiment's error bars)
> WITHOUT EVER referring to, or relying on, synchronising clocks. It only
> assumes a bunch of rulers (identical in construction) and a bunch of
> clocks (identical in construction). The rulers have uniform gauges
> engraved on them identically (when positioned next to each other), the
> clocks tick identically (when positioned next to each other).
>
> We perform the experiments described below far away from any external
> forces (including gravity) so objects when left alone travel uniformly
> in straight lines.
>
> We first establish that the geometry of space is Euclidean. This means
> we need to verify the Pythagorean theorem. Which in turn means we need
> to experimentally set up the model for "right triangle".
>
> This is easily arranged: take two rulers of the same length
> (experimentally verifying beforehand that they are congruent) and
> arrange them in line, end to end. Let's say the first ruler (+++) is AB
> and the second (===) is BC:
>
> ++++++++++==========
> A B C
>
> Now we pick a point D not on that line that's equidistant from A and C.
> Again, "equidistant" can be verified with our rulers. Then we call the
> triangle ABD "right triangle" and we claim the following (Pythagorean
> theorem):
>
> AB^2 + BD^2 = AD^2 (*)
>
> (same thing could be done using the triangle CBD).
>
> If the result (*) holds for any initial configuration of the points A,
> B, D set up according to the above procedure, then we'll have
> established the spatial geometry as Euclidean.
>
> Now comes the less obvious part: establishing the Minkowski geometry
> when time is added as an extra degree of freedom.
>
> This means verifying a "Lorentz" version of the Pythagorean theorem
> which involves two cases: the hypoteneuse being measured by a clock vs.
> the hypoteneuse being measured by a ruler.
>
> Here is our experimental setup:
>
> 1. a light source/detector and a clock (all three co-located). Let's
> call this clock "clock 1".
>
> 2. a ruler with a mirror mounted on it (pointing in the direction along
> the ruler), say at the "0" mark of its gauge engraving. We also put one
> of our clocks at that mirror ("clock 2").
>
> The clocks are not synchronised or offset in any particular way.
>
> We perform two types of experiments.
>
> EXPERIMENT 1 ("the hypoteneuse measured by a clock")
>
> Put the light source somewhere on the ruler and set it in a uniform
> motion along it. Do it so that at some instant the light source and the
> mirror coincide. Take their clock readings at that instant. Let's say
> those readings are C1 and C2. Meaning: clock 1 reads C1, clock 2 reads C2.
>
> Some time after that, let the moving light source emit a light pulse
> toward the mirror. Suppose the light source clock (clock 1) reads C1' at
> the instant of the light pulse emission.
>
> The light travels along the rules and gets reflected by the ruler's
> mirror at its time C2' (clock 2).
>
> The light pulse finally gets back to the light source when its clock
> reads C1''.
>
> We notice that the numbers so collected always obey:
>
> (C1' - C1)(C1'' - C1) = (C2' - C2)^2 (**)
>
> EXPERIMENT 2 ("the hypoteneuse is a ruler")
>
> For this part of the experiment we need to account for the arbitrariness
> of our choice of the rulers' gauge and the clock ticks. We first find
> out what the two-way speed of light is, measured only once, in an
> arbitrarily fixed system (as long as it moves uniformly, as we stipulate
> throughout).
>
> A two-way measurement like this again does not require any clock
> synchronisation and we do not even assume anything special about this:
> no "universality" or "constancy" of this number, no "second postulate",
> none of this. Just measure the two-way speed of light, just once. Call
> the resulting number "V" (in deference to Einstein's 1905 paper).
>
> Now we perform a similar experiment. We begin as in Experiment 1 by
> setting the light source in motion along the ruler. This time we don't
> bother recording the two clocks' readings when the light source and the
> mirror coincide. We just let the process continue.
>
> And again, we emit the light pulse at C1' toward the mirror. Now at some
> arbitrarily chosen instant but BEFORE the reflected light pulse hits the
> light source/detector again, we write down two things concerning clock 1:
>
> - its reading, call it C1*,
> - the ruler's engraved distance marking at that clock position at that
> instant, call it D*.
>
> Then we write down the clock 1 reading when the light pulse hits the
> detector (C1''),
> as before.
>
> We notice that the numbers so collected always obey:
>
> (C1' - C1*)(C1'' - C1*) = -(D*)^2/V^2 (***)
>
> That's it. If the numbers collected in the above experiments satisfy
> (*), (**), and (***), then space and time must necessarily obey the
> entirety of special relativity (time dilation, twin paradox, etc. etc.
> etc.). Of course all that up to the experiment's error bars.
>
> Of course this method of defining special relativity suffers from
> exactly the same shortcomings as the pre-Descartes geometry. But THE
> POINT here is that clock synchronisation is never used or referred to here.
>
> In other words, clock sync is NOT any essential aspect of special
> relativity, let alone a defining one (as it's frequently claimed). Such
> synchronisation, as well as the notion of simultaneity that follows, is
> thus only a convenience for the theory. Just like the choice of
> coordinate axes is only a convenience for Euclidean geometry.
>
> --
> Jan

Two geometries + Conformal Mapping = One Geometry

| |Planck length| x |Planck length| | = root |Planck length|

"Holding station" is synchronization of a sort.

Your "rulers" are usually called "rigid rods".
Your "rulers' engraved distance markings"
are just so many connected rods, or for
Einstein, each their "bridges",
in the "spatial" in GR and the "spacial" in SR.

(That massy bodies move in the spatial
and light-like in the spacial.)

You have just described "Doppler",
which Relativity resolves to in the "classical limit".

What's interesting in quasi-invariant measure theory
about continuum mechanics is the "Doppeler",
doubling and halving spaces, helping explain why
it's so usual the seemingly unjustified "1/2"
shows up so much in usual systems particles/waves.

Then here is "Relativity of Simultaneity is non-local",
yeah, I know that's "not just Restricted a.k.a. Special"
Relativity any more. Yet, it's important for resonance
theory and space contraction and other features of the
Quantum Mechanics in the Quantum Field Theory and the
General Relativity theory's spatial and SR's spacial
all one gauge theory, with supersymmetry, fall gravity,
it's a theory of sum potentials, a fall gravity,
it's a continuum mechanics, ....

So, you should really enjoy Einstein's bridges
as with regards to measure in his
"a total field theory, a differential-system,
of inertial-systems, mass-energy equivalence
in space-contraction, and including the L-principle".

Bonjours

JanPB wrote:

> We perform the experiments described below far away from any external forces
> (including gravity) so objects when left alone travel uniformly in straight
> lines.

Thanks. I had never thought of SR being based purely on
the results of experimental measurement. Beside which,
it appears to me that the direction of SR pedagogy has
been strongly away from reliance on electromagnetic
phenomena (such as the second postulate) for its
derivation. As a "principle theory", the tendency in
many modern treatments of special relativity has been to
base it on the single postulate of Minkowski spacetime,
(as for instance, in textbooks by Taylor and Wheeler or
Callahan) or equivalently, on the single postulate of
universal Lorentz covariance. This is, to a large extent,
a matter of fashion. Whether it is sensible to approach
SR from a purely geometric approach is, as you note,
another topic.

Yes, this is my main beef with the standard modern way of teaching the theory.
It REALLY ought to be taught AFTER a course in electrodynamics which is when
the student can see and understand why such a theory could be thought of
by anyone in the first place. Unfortunately, the very mathematical structure of
the theory allows skipping all that without sacrificing the logic of a derivation.
But this completely kills the students' understanding of the theory, it just
becomes a set of whimsical consequences of some whimsical axioms.

--
Jan

JanPB wrote:

> Yes, this is my main beef with the standard modern way of teaching the theory.
> It REALLY ought to be taught AFTER a course in electrodynamics which is when
> the student can see and understand why such a theory could be thought of
> by anyone in the first place. Unfortunately, the very mathematical structure of
> the theory allows skipping all that without sacrificing the logic of a derivation.
> But this completely kills the students' understanding of the theory, it just
> becomes a set of whimsical consequences of some whimsical axioms.

As you've pointed out on numerous occasions, to Einstein, the
really important part of his paper was the second part which
hardly anybody reads nowadays because his pre-vector analysis
approach to EM is, by modern standards, rather off-putting and
obscure. But it was the reason why he titled the paper "On the
electrodynamics of moving bodies."

For my part, my own main beef with the dominant two-postulate
approach to SR has been the general use of a version of the
second postulate which is far "stronger" than the one that
Einstein employed. Einstein had stated, "Each ray of light
moves in the coordinate system 'at rest' with the definite
velocity V independent of whether this ray of light is
emitted by a body at rest or in motion." This is a completely
natural, easy-to-accept statement.

However, in a rush to get to the "good stuff", modern
statements of the second postulate usually take a shortcut,
asserting something to the effect that "The speed of light
propagating in vacuum is the same in all inertial reference
frames. Any observer in an inertial frame of reference,
regardless of their velocity, will measure that light has a
speed of c, when it propagates in vacuum." Einstein's
easy-to-accept original statement has been replaced by
something deeply counterintuitive.

I believe that many a crackpot has been born from being
presented with a "strong" form of the second postulate.
First impressions matter, and if a student is presented
with a "strong" version of the second postulate, their
deep initial impression may be "this is nonsense!" and
they go through the rest of their lives never letting go
of their initial impression.

ProkaryoticCaspaseHomolog wrote:

> JanPB wrote:

>> Yes, this is my main beef with the standard modern way of teaching the theory.
>> It REALLY ought to be taught AFTER a course in electrodynamics which is when
>> the student can see and understand why such a theory could be thought of
>> by anyone in the first place. Unfortunately, the very mathematical structure of
>> the theory allows skipping all that without sacrificing the logic of a derivation.
>> But this completely kills the students' understanding of the theory, it just
>> becomes a set of whimsical consequences of some whimsical axioms.

> As you've pointed out on numerous occasions, to Einstein, the
> really important part of his paper was the second part which
> hardly anybody reads nowadays because his pre-vector analysis
> approach to EM is, by modern standards, rather off-putting and
> obscure.

This is probably less of an issue, the main reason is that very soon
after the publication of this paper everyone realised (Einstein included)
that it was the first part ("Kinematical") which was really the more
important one as it introduced a new kinematics to classical mechanics which
appeared to be universal. From that POV the second part became "merely" an
example of the application of the new kinematics in the context of
electrodynamics. Originally Einstein thought the first part was just about
making a careful note of some type of measurements which ALONE he showed was
enough to explain the appearance of the Lorentz transformation on physical
grounds.

> But it was the reason why he titled the paper "On the
> electrodynamics of moving bodies."

Yes.

> For my part, my own main beef with the dominant two-postulate
> approach to SR has been the general use of a version of the
> second postulate which is far "stronger" than the one that
> Einstein employed.

Yes, exactly.

> Einstein had stated, "Each ray of light
> moves in the coordinate system 'at rest'

...and only in this one system!

> with the definite
> velocity V independent of whether this ray of light is
> emitted by a body at rest or in motion." This is a completely
> natural, easy-to-accept statement.

Yes. This is just stating Maxwell's theory, basically.

> However, in a rush to get to the "good stuff", modern
> statements of the second postulate usually take a shortcut,
> asserting something to the effect that "The speed of light
> propagating in vacuum is the same in all inertial reference
> frames.

Yes, and this is a completely curved ball out of nowhere (in most
students' minds) with equally bizarre consequences, all of it for
no apparent reason really. That's what always happens when there
is insufficient knowledge: 1. no motivation for the changes, and
2. no understanding or appreciation of the changes and why they
are the way they are (exactly).

Even excellent textbooks like Griffiths' Electrodynamics use this
approach, which is aggravating because at that point he had just
spent *an entire book on introducing electrodynamics*, so he is
perfectly poised to introduce relativity properly. But nooo, out
comes this knee-jerk pedagogy with "just two postulates and just
assume the speed of light is the same in all inertial frames".

Sigh.

--
Jan

On 2/13/24 5:37 AM, JanPB wrote:
> [...]

Note also that for deriving the equations of SR, Einstein's second
postulate can be replaced with any appropriate experimental observation,
such as:
a) pion beams exist that are longer than ~20 meters (both
CERN and Fermilab have had pion beams more than a kilometer
long).
b) No matter how much energy is pumped into a charged particle in
an accelerator, its speed relative to the lab never exceeds c
while its kinetic energy increases correspondingly.
c) The observed trajectories of high-energy charged particles in
magnetic and electric fields, as a function of their kinetic
energy.

Or his second postulate can be replaced with a more abstract,
theoretical statement: there is a finite upper bound on the speed of
information transfer.

(In discussing this last, one should note that it is
inspired by (b). It should also be pointed out that
arbitrarily-large speeds can destroy energy/momentum
conservation, and common notions of causality, because
objects could "zoom in from infinity".)

Tom Roberts

ProkaryoticCaspaseHomolog wrote:

> JanPB wrote:

>> We perform the experiments described below far away from any external forces
>> (including gravity) so objects when left alone travel uniformly in straight
>> lines.

> Thanks. I had never thought of SR being based purely on
> the results of experimental measurement.

And it's not, of course. Poincare was right, you're
an idiot and The Shit is just ma bizarre, violating
common sense convention of speaking.

JanPB wrote:

> In a previous thread ("Einstein's Relativity contains a HUGE Loophole. Its
> Implications Can't Be Ignored") a common assumption is made that clock
> synchronisation is essential to setting up and defining special relativity.

> Then an argument is presented critiquing the relevant clock synchronisation
> (typically, Einstein's) as in some ways deficient or arbitrary or impossible to
> verify experimentally. This then "leads" to the conclusion that special
> relativity is circular at best and self-contradictory at worst.

> But this is an incorrect

And poor idiot Jan is a queen of England too.

, although widespread, assumption. Special relativity
> is a theory which is based on a geometry (whether this is a sensible thing to
> do is another topic) and therefore it does not need coordinate axes, or clock
> sync, for its definition. It's the same with Euclidean geometry: it existed
> just fine before Rene Descartes.

Sure, but your idiot guru had to announce it false,
as it didn't want to fit his madness.

> So I'll describe below the relevant experimental setup that necessarily implies
> special relativity holds (to the experiment's error bars) WITHOUT EVER
> referring to, or relying on, synchronising clocks. It only assumes a bunch of
> rulers (identical in construction) and a bunch of clocks (identical in
> construction).

Sorry, trash, anyone can check GPS, clocks of
identical construction have no mystical power you're
suggesting and are useless in serious measurements.
Too bad for The Shit, too bad for idiots like you.

Tom Roberts wrote:

> On 2/13/24 5:37 AM, JanPB wrote:
>> [...]

> Note also that for deriving the equations of SR, Einstein's second
> postulate can be replaced with any appropriate experimental observation,

And in the meantime in the real world, forbidden by idiots like
you "improper" clocks of GPS or TAI keep measuring y'=t,
just like all serious clocks always did.

Tom Roberts wrote:
>
> Note also that for deriving the equations of SR, Einstein's second
> postulate can be replaced with any appropriate experimental observation,
> such as:
> a) pion beams exist that are longer than ~20 meters (both
> CERN and Fermilab have had pion beams more than a kilometer
> long).
> b) No matter how much energy is pumped into a charged particle in
> an accelerator, its speed relative to the lab never exceeds c
> while its kinetic energy increases correspondingly.
> c) The observed trajectories of high-energy charged particles in
> magnetic and electric fields, as a function of their kinetic
> energy.
>
>
> Or his second postulate can be replaced with a more abstract,
> theoretical statement: there is a finite upper bound on the speed of
> information transfer.

For bradyons. For luxons it's also a lower bound (in vacuum).

> (In discussing this last, one should note that it is
> inspired by (b).

Which are bradyons.

> It should also be pointed out that
> arbitrarily-large speeds can destroy energy/momentum
> conservation, and common notions of causality, because
> objects could "zoom in from infinity".)
>
> Tom Roberts

I don't think so, Tom. Bilaniuk et al. in "Meta relativity"
showed that FTL particles would obey E = |m|c^2/sqrt(u^2/c^2 - 1),
which approaches zero as u approaches infinity. Of course
they could still come in from space-like distances. Even the
LT allows that: u' = (u - v)/(1 - uv/c^2), where u < c^2/v.

The problem, of course, is the denominator which is because of

dt' = gamma(dt - v*dx/c^2), which has the same proscribed domain.

The argument that the domain for v1 > v rules out the extended
domain for v is false, however.

Gary

On 02/13/2024 03:37 AM, JanPB wrote:
> ProkaryoticCaspaseHomolog wrote:
>
>> JanPB wrote:
>
>>> Yes, this is my main beef with the standard modern way of teaching
>>> the theory.
>>> It REALLY ought to be taught AFTER a course in electrodynamics which
>>> is when
>>> the student can see and understand why such a theory could be thought of
>>> by anyone in the first place. Unfortunately, the very mathematical
>>> structure of
>>> the theory allows skipping all that without sacrificing the logic of
>>> a derivation.
>>> But this completely kills the students' understanding of the theory,
>>> it just
>>> becomes a set of whimsical consequences of some whimsical axioms.
>
>> As you've pointed out on numerous occasions, to Einstein, the
>> really important part of his paper was the second part which
>> hardly anybody reads nowadays because his pre-vector analysis approach
>> to EM is, by modern standards, rather off-putting and
>> obscure.
>
> This is probably less of an issue, the main reason is that very soon
> after the publication of this paper everyone realised (Einstein included)
> that it was the first part ("Kinematical") which was really the more
> important one as it introduced a new kinematics to classical mechanics
> which
> appeared to be universal. From that POV the second part became "merely" an
> example of the application of the new kinematics in the context of
> electrodynamics. Originally Einstein thought the first part was just about
> making a careful note of some type of measurements which ALONE he showed
> was
> enough to explain the appearance of the Lorentz transformation on physical
> grounds.
>
>> But it was the reason why he titled the paper "On the
>> electrodynamics of moving bodies."
>
> Yes.
>
>> For my part, my own main beef with the dominant two-postulate approach
>> to SR has been the general use of a version of the
>> second postulate which is far "stronger" than the one that
>> Einstein employed.
>
> Yes, exactly.
>
>> Einstein had stated, "Each ray of light moves in the coordinate system
>> 'at rest'
>
> ..and only in this one system!
>
>> with the definite velocity V independent of whether this ray of light
>> is emitted by a body at rest or in motion." This is a completely
>> natural, easy-to-accept statement.
>
> Yes. This is just stating Maxwell's theory, basically.
>
>> However, in a rush to get to the "good stuff", modern statements of
>> the second postulate usually take a shortcut,
>> asserting something to the effect that "The speed of light propagating
>> in vacuum is the same in all inertial reference frames.
>
> Yes, and this is a completely curved ball out of nowhere (in most
> students' minds) with equally bizarre consequences, all of it for
> no apparent reason really. That's what always happens when there
> is insufficient knowledge: 1. no motivation for the changes, and
> 2. no understanding or appreciation of the changes and why they
> are the way they are (exactly).
>
> Even excellent textbooks like Griffiths' Electrodynamics use this
> approach, which is aggravating because at that point he had just
> spent *an entire book on introducing electrodynamics*, so he is
> perfectly poised to introduce relativity properly. But nooo, out
> comes this knee-jerk pedagogy with "just two postulates and just
> assume the speed of light is the same in all inertial frames".
>
> Sigh.
>
> --
> Jan

These days sometimes it's that the

electrodynamics of moving bodies

is framed in the

electrodynamics of extended objects.