• Optimum Track

    From David Entwistle@qnivq.ragjvfgyr@ogvagrearg.pbz to rec.puzzles on Mon May 4 01:00:16 2026
    From Newsgroup: rec.puzzles

    A fair time ago (perhaps 50 years) I remember reading a question in a
    physics text book. I may have forgotten some detail, but I think it went
    as follows:

    You are asked to construct a track which will allow a marble to travel,
    under the influence of gravity, between two points in a vertical plane. If
    the start point (A) is at point (0, 1) and the end point (B) is at point
    (1, 0) what is the nature of the curve of the track that allows the
    journey to be completed in the minimum time? You can ignore friction and
    air resistance.

    I remember the answer given. Does anyone have any pointers how you may
    tackle this question?

    Solution provided:

    N plpybvq...
    --
    David Entwistle
    --- Synchronet 3.21f-Linux NewsLink 1.2
  • From Ilan Mayer@user4643@newsgrouper.org.invalid to rec.puzzles on Mon May 4 02:32:33 2026
    From Newsgroup: rec.puzzles


    David Entwistle <qnivq.ragjvfgyr@ogvagrearg.pbz> posted:

    A fair time ago (perhaps 50 years) I remember reading a question in a physics text book. I may have forgotten some detail, but I think it went
    as follows:

    You are asked to construct a track which will allow a marble to travel, under the influence of gravity, between two points in a vertical plane. If the start point (A) is at point (0, 1) and the end point (B) is at point
    (1, 0) what is the nature of the curve of the track that allows the
    journey to be completed in the minimum time? You can ignore friction and
    air resistance.

    I remember the answer given. Does anyone have any pointers how you may tackle this question?

    Solution provided:

    N plpybvq...


    https://en.wikipedia.org/wiki/Brachistochrone_curve
    --- Synchronet 3.21f-Linux NewsLink 1.2
  • From Mike Terry@news.dead.person.stones@darjeeling.plus.com to rec.puzzles on Mon May 4 04:05:10 2026
    From Newsgroup: rec.puzzles

    On 04/05/2026 02:00, David Entwistle wrote:
    A fair time ago (perhaps 50 years) I remember reading a question in a
    physics text book. I may have forgotten some detail, but I think it went
    as follows:

    You are asked to construct a track which will allow a marble to travel,
    under the influence of gravity, between two points in a vertical plane. If the start point (A) is at point (0, 1) and the end point (B) is at point
    (1, 0) what is the nature of the curve of the track that allows the
    journey to be completed in the minimum time? You can ignore friction and
    air resistance.

    I remember the answer given. Does anyone have any pointers how you may
    tackle this question?

    Solution provided:

    N plpybvq...


    I think you're remembering The Brachistochrone Problem, which can be solved using "Calculus of
    Variations". There are (e.g.) YouTube videos explaining this, and I doubt I could give a better
    explanation myself!

    The Brachistochrone problem is slightly different from your description - it uses a point particle,
    sliding frictionlessly on a track (or wire frame). When you change it to a marble rolling on a
    track you introduce an angular rotation component (the rotation of marble) which will distort the
    solution so I would expect the solution to be different...


    Mike.

    --- Synchronet 3.21f-Linux NewsLink 1.2
  • From James Dow Allen@user4353@newsgrouper.org.invalid to rec.puzzles on Mon May 4 18:42:55 2026
    From Newsgroup: rec.puzzles


    Mike Terry <news.dead.person.stones@darjeeling.plus.com> posted:

    I think you're remembering The Brachistochrone Problem, which can be solved using "Calculus of
    Variations". There are (e.g.) YouTube videos explaining this, and I doubt I could give a better
    explanation myself!

    The Brachistochrone problem is slightly different from your description - it uses a point particle,
    sliding frictionlessly on a track (or wire frame). When you change it to a marble rolling on a
    track you introduce an angular rotation component (the rotation of marble) which will distort the
    solution so I would expect the solution to be different...


    Good point.

    A co-worker achieved some of his fame by solving the
    "unrestrained brachistochrone" problem. Google will lead you to
    the problem statement, along with abstracts of that man's papers on
    the topic, though the content may be paywalled.

    35 years ago he challenged me with that problem when I stumbled into his
    office to say hello; he was impressed when I quickly intuited the solution. Unfortunately the apogee of my math skill is now well in the past.... :-{

    FWIW, he failed when I challenged him back with
    "Find the fallacy in this 'proof' of the Four-Color Map Theorem" --
    https://fabpedigree.com/james/color4.htm

    Mike.

    James
    --- Synchronet 3.21f-Linux NewsLink 1.2
  • From Charlie Roberts@croberts@gmail.com to rec.puzzles on Mon May 4 16:58:54 2026
    From Newsgroup: rec.puzzles

    On Mon, 04 May 2026 18:42:55 GMT, James Dow Allen <user4353@newsgrouper.org.invalid> wrote:

    A co-worker achieved some of his fame by solving the
    "unrestrained brachistochrone" problem. Google will lead you to
    the problem statement, along with abstracts of that man's papers on
    the topic, though the content may be paywalled.

    35 years ago he challenged me with that problem when I stumbled into his >office to say hello; he was impressed when I quickly intuited the solution. >Unfortunately the apogee of my math skill is now well in the past.... :-{

    Can you tell us who the person is? Or, is it classified?
    --- Synchronet 3.21f-Linux NewsLink 1.2
  • From Charlie Roberts@croberts@gmail.com to rec.puzzles on Mon May 4 16:53:53 2026
    From Newsgroup: rec.puzzles

    On Mon, 4 May 2026 04:05:10 +0100, Mike Terry <news.dead.person.stones@darjeeling.plus.com> wrote:

    On 04/05/2026 02:00, David Entwistle wrote:
    A fair time ago (perhaps 50 years) I remember reading a question in a
    physics text book. I may have forgotten some detail, but I think it went
    as follows:

    You are asked to construct a track which will allow a marble to travel,
    under the influence of gravity, between two points in a vertical plane. If >> the start point (A) is at point (0, 1) and the end point (B) is at point
    (1, 0) what is the nature of the curve of the track that allows the
    journey to be completed in the minimum time? You can ignore friction and
    air resistance.

    ......


    I think you're remembering The Brachistochrone Problem, which can be solved using "Calculus of
    Variations". There are (e.g.) YouTube videos explaining this, and I doubt I could give a better
    explanation myself!

    The Brachistochrone problem is slightly different from your description - it uses a point particle,
    sliding frictionlessly on a track (or wire frame). When you change it to a marble rolling on a
    track you introduce an angular rotation component (the rotation of marble) which will distort the
    solution so I would expect the solution to be different...

    We can give David a break as he said that his marble "travels", and
    not "rolls" .... it would take some friction to get the marble rolling
    if it initially not spinning, though we can really get caught up here
    in things like the transition from sliding to rolling and all that
    (which requires friction, anyway, to make the transition).

    As a lapsed theoretical physicist, I always thought that the way
    to solve the (restricted) problem via the traditional calculus
    of variations route. A few years back I stumbled across this
    excellent video by 3Blue1Brown that shows how it can be done
    by adapting optics, in particular, Snell's Law and Fermat's principle
    of least time. You will find it here:

    https://www.youtube.com/watch?v=Cld0p3a43fU

    (... which reminds me that I should get hold of Levi's
    book!)

    A more formal paper on this is "BernoulliAs light ray solution
    of the brachistochrone problem through HamiltonAs eyes"
    by Henk W. Broer. You can find it at

    www.math.rug.nl/~broer/pdf/ws-ijbc.pdf

    This got me interested in looking for a similar side door
    to the unrestrained problem. I have not found one, but
    this paper is something worth going through:

    https://www.researchgate.net/publication/348379398_The_unrestrained_brachistochrone_with_sliding_friction


    --- Synchronet 3.21f-Linux NewsLink 1.2
  • From James Dow Allen@user4353@newsgrouper.org.invalid to rec.puzzles on Tue May 5 05:11:21 2026
    From Newsgroup: rec.puzzles


    Charlie Roberts <croberts@gmail.com> posted:

    Can you tell us who the person is? Or, is it classified?

    ???

    (1) The suggested Googling would have revealed his identity in
    less time than it took you to post this.
    (2) Is it not rude, at least in many or most contexts, to post
    unnecessarily the names of private individuals?

    James

    --- Synchronet 3.21f-Linux NewsLink 1.2
  • From James Dow Allen@user4353@newsgrouper.org.invalid to rec.puzzles on Tue May 5 05:23:38 2026
    From Newsgroup: rec.puzzles


    James Dow Allen <user4353@newsgrouper.org.invalid> posted:

    (1) The suggested Googling would have revealed his identity in
    less time than it took you to post this.
    (2) Is it not rude, at least in many or most contexts, to post
    unnecessarily the names of private individuals?

    (3) No, (1) and (2) are not contradictory. The identity IS readily available to anyone with need or desire to know. But I do not need or desire to
    PUBLISH the name and facilitate data-harvesting engines to link it to my anecdote about a fallacious proof.

    James
    --- Synchronet 3.21f-Linux NewsLink 1.2
  • From David Entwistle@qnivq.ragjvfgyr@ogvagrearg.pbz to rec.puzzles on Tue May 5 14:50:16 2026
    From Newsgroup: rec.puzzles

    On Mon, 04 May 2026 02:32:33 GMT, Ilan Mayer wrote:

    https://en.wikipedia.org/wiki/Brachistochrone_curve

    Thanks that's very interesting.
    --
    David Entwistle
    --- Synchronet 3.21f-Linux NewsLink 1.2
  • From David Entwistle@qnivq.ragjvfgyr@ogvagrearg.pbz to rec.puzzles on Tue May 5 14:59:03 2026
    From Newsgroup: rec.puzzles

    On Mon, 4 May 2026 04:05:10 +0100, Mike Terry wrote:

    I think you're remembering The Brachistochrone Problem, which can be
    solved using "Calculus of Variations". There are (e.g.) YouTube videos explaining this, and I doubt I could give a better explanation myself!

    Thanks. I watched the 3Brown1Blue video 'The Brachistochrone, with Steven Strogatz'. As always, with 3B1B videos, it was excellent.

    https://www.youtube.com/watch?v=Cld0p3a43fU


    The Brachistochrone problem is slightly different from your description

    I'm sure that is just my defective recollection of the original problem.
    --
    David Entwistle
    --- Synchronet 3.21f-Linux NewsLink 1.2
  • From Mike Terry@news.dead.person.stones@darjeeling.plus.com to rec.puzzles on Sun May 10 04:29:28 2026
    From Newsgroup: rec.puzzles

    On 04/05/2026 19:42, James Dow Allen wrote:

    Mike Terry <news.dead.person.stones@darjeeling.plus.com> posted:

    I think you're remembering The Brachistochrone Problem, which can be solved using "Calculus of
    Variations". There are (e.g.) YouTube videos explaining this, and I doubt I could give a better
    explanation myself!

    The Brachistochrone problem is slightly different from your description - it uses a point particle,
    sliding frictionlessly on a track (or wire frame). When you change it to a marble rolling on a
    track you introduce an angular rotation component (the rotation of marble) which will distort the
    solution so I would expect the solution to be different...


    Good point.

    A co-worker achieved some of his fame by solving the
    "unrestrained brachistochrone" problem. Google will lead you to
    the problem statement, along with abstracts of that man's papers on
    the topic, though the content may be paywalled.

    35 years ago he challenged me with that problem when I stumbled into his office to say hello; he was impressed when I quickly intuited the solution. Unfortunately the apogee of my math skill is now well in the past.... :-{

    FWIW, he failed when I challenged him back with
    "Find the fallacy in this 'proof' of the Four-Color Map Theorem" --
    https://fabpedigree.com/james/color4.htm

    Hmmm, I like the "proof", because as presented it sounds plausible, and is "nearly" correct. I
    mean, all of the logic is correct apart from one (crucial) step!

    I'm not sure if this is a puzzle, but just in case I'll put in some spoiler space...


    .s......
    ..p.....
    ..o.....
    .i......
    ..l.....
    ...e....
    ....r...
    .....s..
    .....p..
    ....o...
    .....i..
    ......l.
    ......r.
    .....r..
    ....s...
    ...p....
    ..o.....
    ..i.....
    ...l....
    ..e.....
    .r......
    .s......
    ..p.....
    ..o.....
    .i......
    ..l.....
    ...e....
    ....r...
    .....s..
    .....p..
    ....o...
    .....i..
    ......l.
    ......r.
    .....r..
    ....s...
    ...p....
    ..o.....
    ..i.....
    ...l....
    ..e.....
    .r......


    [View with fixed font, of course!]

    So here is a counter-example for the (fallacious) reasoning presented in your link

    ---------------1---------------
    | / \ |
    | / \ |
    | / \ |
    | / \ |
    | 2 2 |
    | /| |\ |
    | / | | \ |
    | / | | \ |
    | / | | \ |
    | / 3---------4 \ |
    | / \ / \ |
    | / \ / \ |
    | / \ / \ |
    |/ \ / \|
    4--------------1--------------3
    \ /
    \ /
    \ /
    -----------------------

    The numbers above represent the vertices (nodes) of the graph:
    1 = green
    2 = red
    3 = blue
    4 = yellow
    The ascii lines are the edges. The pentagon (1,2,3,4,2) starting at the top 1 is the main working
    pentagon of the (fallacious) proof. (The temporarily deleted node inside the pentagon is not shown.
    Also the pentagon is rotated a bit from the proof, but it will make sense if you've followed the
    proof.)

    The presented proof claims (correctly) that we can change the leftmost 2 to a 4 (and proceeding with
    flipping 2<-->4 throughout the associated (2,4)-connected-subgraph). I.e. specifically in my graph
    we can change the leftmost 2 to a 4 and the rightmost 4 to a 2, and the colouring is still valid.
    [but the pentagon still has 4 colours!]

    Alternatively, /starting from the original graph/ we can perform the similar flipping 2<-->3 of the
    rightmost 2 and 3, and the colouring remains valid. [but of course the pentagon still has 4 colours!]

    The presented proof suggests that by doing /both/ of these flips together, the pentagon would
    contain only 3 colours [correct], and the new colouring must still be valid [INCORRECT]. As seen in
    my graph above, if we apply both flipping operations, the bottom 3 and 4 nodes both become 2 (red)
    and are adjacent, so the colouring is broken.


    The presented proof chooses to draw its 1-3 and 1-4 "fences" in a way which isolates the regions
    they enclose, making the false conclusion sound plausible - it works if the fences are indeed
    isolated as the presented proof draws them. However, the two fences may cross each other at a
    1-node, as shown in my example, and then the implied isolation of the to-be-flipped (2,4)- and
    (2,3)- subgraphs no longer holds [..so they cannot both be performed together without breaking the
    colouring].

    In my example graph, the 1-3 and 1-4 fences cross each other at the bottom 1-node.

    Mike.

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Mike Terry@news.dead.person.stones@darjeeling.plus.com to rec.puzzles on Sun May 10 16:58:12 2026
    From Newsgroup: rec.puzzles

    On 10/05/2026 04:29, Mike Terry wrote:
    On 04/05/2026 19:42, James Dow Allen wrote:

    Mike Terry <news.dead.person.stones@darjeeling.plus.com> posted:

    I think you're remembering The Brachistochrone Problem, which can be solved using "Calculus of
    Variations".a There are (e.g.) YouTube videos explaining this, and I doubt I could give a better
    explanation myself!

    The Brachistochrone problem is slightly different from your description - it uses a point particle,
    sliding frictionlessly on a track (or wire frame).a When you change it to a marble rolling on a
    track you introduce an angular rotation component (the rotation of marble) which will distort the
    solution so I would expect the solution to be different...


    Good point.

    A co-worker achieved some of his fame by solving the
    "unrestrained brachistochrone" problem.a Google will lead you to
    the problem statement, along with abstracts of that man's papers on
    the topic, though the content may be paywalled.

    35 years ago he challenged me with that problem when I stumbled into his
    office to say hello; he was impressed when I quickly intuited the solution. >> Unfortunately the apogee of my math skill is now well in the past.... :-{

    FWIW, he failed when I challenged him back with
    "Find the fallacy in this 'proof' of the Four-Color Map Theorem" --
    aaaa https://fabpedigree.com/james/color4.htm

    Hmmm,a I like the "proof", because as presented it sounds plausible, and is "nearly" correct.a I
    mean, all of the logic is correct apart from one (crucial) step!

    I'm not sure if this is a puzzle, but just in case I'll put in some spoiler space...


    .s......
    ..p.....
    ..o.....
    .i......
    ..l.....
    ...e....
    ....r...
    .....s..
    .....p..
    ....o...
    .....i..
    ......l.
    ......r.
    .....r..
    ....s...
    ...p....
    ..o.....
    ..i.....
    ...l....
    ..e.....
    .r......
    .s......
    ..p.....
    ..o.....
    .i......
    ..l.....
    ...e....
    ....r...
    .....s..
    .....p..
    ....o...
    .....i..
    ......l.
    ......r.
    .....r..
    ....s...
    ...p....
    ..o.....
    ..i.....
    ...l....
    ..e.....
    .r......


    [View with fixed font, of course!]

    So here is a counter-example for the (fallacious) reasoning presented in your link

    aa ---------------1---------------
    aa |aaaaaaaaaaaa / \aaaaaaaaaaaa |
    aa |aaaaaaaaaaa /aa \aaaaaaaaaaa |
    aa |aaaaaaaaaa /aaaa \aaaaaaaaaa |
    aa |aaaaaaaaa /aaaaaa \aaaaaaaaa |
    aa |aaaaaaaa 2aaaaaaaa 2aaaaaaaa |
    aa |aaaaaaa /|aaaaaaaa |\aaaaaaa |
    aa |aaaaaa / |aaaaaaaa | \aaaaaa |
    aa |aaaaa /a |aaaaaaaa |a \aaaaa |
    aa |aaaa /aa |aaaaaaaa |aa \aaaa |
    aa |aaa /aaa 3---------4aaa \aaa |
    aa |aa /aaaaa \aaaaaa /aaaaa \aa |
    aa |a /aaaaaaa \aaaa /aaaaaaa \a |
    aa | /aaaaaaaaa \aa /aaaaaaaaa \ |
    aa |/aaaaaaaaaaa \ /aaaaaaaaaaa \|
    aa 4--------------1--------------3
    aaa \aaaaaaaaaaaaaaaaaaaaaaaaaa /
    aaaa \aaaaaaaaaaaaaaaaaaaaaaaa /
    aaaaa \aaaaaaaaaaaaaaaaaaaaaa /
    aaaaaa -----------------------

    The numbers above represent the vertices (nodes) of the graph:
    a 1 = green
    a 2 = red
    a 3 = blue
    a 4 = yellow
    The ascii lines are the edges.a The pentagon (1,2,3,4,2) starting at the top 1 is the main working
    pentagon of the (fallacious) proof.a (The temporarily deleted node inside the pentagon is not shown.
    aAlso the pentagon is rotated a bit from the proof, but it will make sense if you've followed the
    proof.)

    The presented proof claims (correctly) that we can change the leftmost 2 to a 4 (and proceeding with
    flipping 2<-->4 throughout the associated (2,4)-connected-subgraph).a I.e. specifically in my graph
    we can change the leftmost 2 to a 4 and the rightmost 4 to a 2, and the colouring is still valid.
    [but the pentagon still has 4 colours!]

    Sorry, that should have been "...we can change the leftmost 2 to a 4 and the /leftmost/ 4 to a 2.."


    Alternatively, /starting from the original graph/ we can perform the similar flipping 2<-->3 of the
    rightmost 2 and 3, and the colouring remains valid.a [but of course the pentagon still has 4 colours!]

    The presented proof suggests that by doing /both/ of these flips together, the pentagon would
    contain only 3 colours [correct], and the new colouring must still be valid [INCORRECT].a As seen in
    my graph above, if we apply both flipping operations, the bottom 3 and 4 nodes both become 2 (red)
    and are adjacent, so the colouring is broken.


    The presented proof chooses to draw its 1-3 and 1-4 "fences" in a way which isolates the regions
    they enclose, making the false conclusion sound plausible - it works if the fences are indeed
    isolated as the presented proof draws them.a However, the two fences may cross each other at a
    1-node, as shown in my example, and then the implied isolation of the to-be-flipped (2,4)- and
    (2,3)- subgraphs no longer holds [..so they cannot both be performed together without breaking the
    colouring].

    In my example graph, the 1-3 and 1-4 fences cross each other at the bottom 1-node.

    Mike.

    --- Synchronet 3.22a-Linux NewsLink 1.2