Path: ...!2.eu.feeder.erje.net!feeder.erje.net!feeder1-2.proxad.net!proxad.net!feeder1-1.proxad.net!cleanfeed3-a.proxad.net!nnrp1-1.free.fr!not-for-mail
Newsgroups: sci.lang,alt.usage.english
Subject: Re: OT: Converting miles/km
From: nospam@de-ster.demon.nl (J. J. Lodder)
Reply-To: jjlxa31@xs4all.nl (J. J. Lodder)
Date: Thu, 26 Sep 2024 11:56:41 +0200
References: <slrnvepbvk.tfc.naddy@lorvorc.mips.inka.de> <vd1cdj$3nvqo$1@dont-email.me> <66f476b2$0$1308$426a74cc@news.free.fr> <vd2n1k$23kh$1@dont-email.me>
Organization: De Ster
Mail-Copies-To: nobody
User-Agent: MacSOUP/2.8.5 (ea919cf118) (Mac OS 10.12.6)
Lines: 168
Message-ID: <66f52fd9$0$22677$426a74cc@news.free.fr>
NNTP-Posting-Date: 26 Sep 2024 11:56:41 CEST
NNTP-Posting-Host: 213.10.137.58
X-Trace: 1727344601 news-2.free.fr 22677 213.10.137.58:49654
X-Complaints-To: abuse@proxad.net
Bytes: 8743

Jeff Barnett <jbb@notatt.com> wrote:

> On 9/25/2024 2:46 PM, J. J. Lodder wrote:
> > Jeff Barnett <jbb@notatt.com> wrote:
> > 
> >> On 9/19/2024 5:12 PM, Christian Weisgerber wrote:
> >>> I'm sorry, I don't know where to post this.  I'm crossposting to
> >>> alt.usage.english, because statute miles as a unit mostly afflict
> >>> the English-speaking world.
> >>>
> >>> So you want to convert between miles and kilometers.  The conversion
> >>> factor is... uh...  A 40-year-old calculator book provides a useful
> >>> tip:  Unless you're designing a space probe, you can use ln(5).
> >>>
> >>> WHAT?
> >>>
> >>> Yes, the natural logrithm of 5 approximates the conversion factor
> >>> between miles and kilometers; specifically one mile is about ln(5)
> >>> kilometers.  It's accurate to four digits.
> >>>
> >>> If nothing else, it's faster to type on a calculator.
> >>>
> >>> I think that's hysterical.
> >>>
> >>
> >>
> >> After glancing at the discussion that follows this post, I thought it
> >> appropriate to point out the book "Dimensional Analysis" New Haven: Yale
> >> University Press (1922) by the Nobel Prize winning physicist Percy
> >> Williams Bridgman.
> > 
> > He didn't win the nobel for this book.
> > (or for his peculiar philosophy of sciece)
> > It is one of those books that many know exists,
> > but few will actually have seen it, let alone read any of it.
> > (don't worry, no loss)
> > You will need a good old university library to find it,
> > or you may find a very rare antiquarian copy,
> > or an almost as rare and by now also antiquarian reprint.
> > 
> >> It essentially describes and defines physical
> >> dimensions such as distance, speed, energy, force, etc. as well as units
> >> that are defined within a dimension such as meters, feet, and microns as
> >> distances. It shows that dimensions MUST match on both sides of an
> >> equation and, if not, there must be multiplicative constants that have
> >> appropriate dimensions to restore balance. You may define base
> >> dimensions and the others in terms of the base. For example, length,
> >> mass, and time to do mechanics.
> > 
> > All completely trivial.
> > What's more, the subject matter has been almost completely forgotten.
> > All that remains is elementary high school knowledge
> > of the -conventional- systems of dimensions
> > that is nowadays associated with the SI.
> > Few people even know anymore that other systems of dimensions
> > are possible.
> > The misconception that a 'dimension' is somehow a property
> > of a physical quantity is shared nearly universally.
> > 
> >> Within an equation, you must use the same units everyplace for
> >> quantities in a specific dimension or dimensionless units of conversion
> >> such as 12 inches per foot. It even shows how to determine when physics
> >> equations express nonsense because of unit disparity or non matching
> >> dimensions.
> > 
> > You may crash Mars landers through non-matching units,
> > never by non-matching dimensions.
> > 
> >> The cherry on the cake is discovery of new physical laws via
> >> dimensional analysis.
> > 
> > Not really. At best it allows you to guess at the form.
> > The book codifies the obvious.
> > Dimensional analysis was already well known and understood
> > through the works of the 19th century greats, such as Kelvin
> > and Rayleigh.
> > The use of the so called 'dimensionless numbers',
> > such as Reynolds', or Froude's number was already well established.
> > 
> >> If you can obtain access to a copy of this book, I recommend taking a
> >> spin through it.
> > 
> > A waste of time and perhaps also money, if you don't mind me saying so.
> > 
> >> A hundred years ago it was novel and educated some very
> >> bright individuals who hadn't quite caught on to what your current
> >> discussion is all about. It wasn't all that obvious way back when. Of
> >> course it was as soon as the subject was systematically presented.
> > 
> > Already then, Bridgman was belabouring the obvious,
> 
> Let me start by pointing out that I don't believe I implied that his 
> Noble Prize was for this book; I know it wasn't. I'm assuming from the
> above that you haven't read the book. There is material in it that you
> must have skipped or don't remember if you had.

That is quite possible. I have held a copy in my hands long ago,
and glanced at some of the pages. Enough of them to justify my opinion
expressed here that he is belabouring the obvious with to many examples,
while not expression the general idea of 'dimension' clearly enough.
It may have been a first edition.

> By the way, much to my 
> surprise new paperback copies are available from Amazon for a modest 
> price. The copy I have was made on a xerox machine 50+ years ago and is
> torture to read - every page has a different slant.
> 
> It's mostly true that unit mistakes can cause mayhem but so can 
> dimensional mistakes. I remember helping track down a calculation that
> would not balance on Apollo because a newbie engineer didn't realize 
> that "knots" measured speed, not distance. No harm was done, just a big
> waste of time. This is all apropos of the discussion in this thread.

Certainly, no need to inform me about the such problems.
I have wasted some time correcting someone who had used old data,
assuming that 'statvolt' is just an oldfashioned name for 'volt'.
(this is both a unit error and a dimension error)
 
> There were other similar mistakes that were common - I suppose these 
> events (as well as greed) were responsible for the huge people and 
> project redundancies on Apollo.
> 
> Later on in a different world, I invented some unit and dimension 
> software for the Symbolics Lisp Machines. Dimensions had the status of
> data types and units were presentation types. So if you wanted to input,
> say an energy, the mouse would highlight and retrieve both erg and joule
> values but not numbers that represented forces. The developer could 
> define the system (e.g. mks or fps) by preferred units as well as the
> electrostatic system in play. Conversions were automatic and, assuming
> the programmer didn't bitch things up, neither could his users. I 
> mentioned this stuff to a physicist and was told to go read his copy of
> Bridgman.

I know.
This is automating those crazy American engineering formulae like:

X/[unit of X] = [strange dimensioned constant] times [A / unit of A to
be used] times [B / unit for B] .....  divided by [F/ unit of F]

> The book is chuck full of examples that show world class physicist 
> making unit and dimension errors. It also shows some techniques to avoid
> them. And as I said above some new physics was discovered using the 
> described techniques.

Possibly, but I wouldn't know an example of new physics -derived- in
this way.
All I know about are swindles, rederiving what you know already. [1]

> Once again, I remind you the book is 100 years 
> old. The above discussion shows that folks in the arts and sciences are
> still making elementary mistakes of the sorts described.

Of course, there is no end to the mistakes people will make.
By Murphy, the best remedy is not correction afterwards,
but using sane systems beforehand
that makes most of those errors impossible to make.

Jan

-- 
"Don't tell me what is wrong with the works of others. Do something good
yourself!" (a grumpy old professor I knew long ago)

[1] Typical example: It is sometimes claimed that you can -derive-
the period of a pendulum using dimensional arguments,
as depending on length of the pendulum and small g only.
(and hence independent of amplitude)
This is a typical example of assuming what needs to be shown.