deathanatos 17 hours ago

> Earth rotates around its axis – one rotation is called a day

A [solar] day is the time between noons, which is slightly more than one rotation on our axis. A single rotation is a [sidereal day] — the Wikipedia article has a good animation.

(The ellipse part of our orbit means the length of a solar day isn't consistent, as the rotation required to get back to pointing at the sun isn't the same throughout the year, which is what leads to mean solar time. The the article doesn't want to do ellipse orbits, which is fine… for a moment… but…)

> When the tick comes directly under the sun, that's (solar) noon, and one full rotation is one day.

But this is sort of where if you do you MST (if you have a fixed day length, you are), then when the tick is directly under the sun it won't necessarily be noon. The difference (between MST & solar) is like 17 minutes at its peak. The aliens will be looking at this going "it's uh… close? But off."

I still think "solar time" is a cultural assumption, though I do think there's a high likelihood of aliens sharing that assumption. But one might also imagine a species on a tidally locked world (maybe having grown up in the twilight region) with no concept of "days".

[sidereal day]: https://en.wikipedia.org/wiki/Sidereal_time

  • senko 14 hours ago

    Thanks for the detailed explanation, I love this - exactly the sort of "geek out" mode I was in when tinkering with the clock.

    I had a fuzzy sense of these differences but had no idea that MST/solar is 17 minutes off - that's a lot! Of course there's also the difference between this and proper clock time (depending on where in your timezone you're located), and the clock shows clock time.

    • deathanatos 4 hours ago

      > but had no idea that MST/solar is 17 minutes off - that's a lot!

      [This Wikipedia article](https://en.wikipedia.org/wiki/Solar_time#Mean_solar_time) has what I think is a good intuition here:

      > Long or short days occur in succession, so the difference builds up until mean time is ahead of apparent time by about 14 minutes near February 6, and behind apparent time by about 16 minutes near November 3.

      With the really nitty gritty being here: https://en.wikipedia.org/wiki/Equation_of_time#Concept

      > Of course there's also the difference between this and proper clock time (depending on where in your timezone you're located), and the clock shows clock time.

      "Standard time", yes, will be different yet. I didn't comment on that b/c the whole article is, of course, building a clock that unmoors us from cultural conventions, and I think Standard Time falls very squarely in the "cultural convention" bucket. So it's more of a solar clock in that sense.

  • TeMPOraL 11 hours ago

    I regret not learning about this before, but apparently "sidereal" is from Latin, and not what I always assumed, i.e. "side real" as in "kinda not quite real, wtf?!" day.

amarant 15 hours ago

After staring at it some more, I'd suggest perhaps switching places between the day of month and month of year circles. That way you would have consistently smaller unit of time=smaller circle.

  • senko 14 hours ago

    That makes sense. I'm least happy with the (day-in-)month circle, but felt the calendar portion wasn't really useful without handling it somehow.

    I wanted the year circle to evoke Earth's orbit, so it had to be near the Sun. Inserting another circle was breaking that intuition (already stretched, tbh) for me. I was flip-flopping between these two and finally decided on the current solution.

    I'm even more unhappy[0] with the fact that the number of ticks and speed on that circle changes every month, depending on the calendar month duration. For a world clock, on month changes it also means that you can't represent the date correctly for all shown timezones.

    I could have used some real-world phenomena that are close to our calendar month instead[1], as pointed out by another commenter. This would make it "more correct" but "less useful in everyday situations". Even though I don't plan to use it every day, one of the goals of the project was for it to aspire to being useful.

    [0] unhappy is a strong word; I mean I'm not satisfied with the solution, but was unable to find one I feel better about [1] https://en.wikipedia.org/wiki/Lunar_month#Types

  • jgtrosh 15 hours ago

    Yeah, this screamed US date format to me! (Though there's no year in this clock)

    • amarant 12 hours ago

      Ha! I had that thought too! hhMMdd is how Americans do it, right?

throwawayk7h 4 hours ago

Meant to be free of cultural preconceptions, but it uses 24-hour division.

There might be a natural, Schelling division to use -- the same ratio year:day could be applied, day:minute. Thus, the day can be broken up into 365.25 "Schelling-minutes" (corresponds to 3.94 SI minutes), and further into 365.25^2 "Schelling-seconds" (corresponds to 0.65 SI seconds). I think it's interesting that we get something not too distant from our own minutes and seconds if we do this.

Of course, for practical daily usage, lacking anything similar to an hour would be a problem. A "half-division" might be in order; that is, √365.25, or roughly 19.11 divisions of the day. Sure, that's close enough to SI hours... but it's probably worth remembering that there's a reason we don't like to use fractional divisions. :P

  • senko 3 hours ago

    Yeah, as @kogold mentioned in another comment - turns out if you want to make a clock that's usable in daily life, you have to import some cultural norms.

vunderba 16 hours ago

Very cool. Kind of reminds me a bit of my solar system clock though I did mine using ThreeJS.

https://clocks.specr.net/solar-system

Senko you might also be interested in this one - it doesn't go into the same depth as yours but it's visually similar.

https://www.clockfaceonline.co.uk/clocks/circles/

  • senko 14 hours ago

    These are nice, thanks for sharing!

    I love your reinterpretation of time as a mini solar system. On impulse I tried to drag to rotate it and it worked, was delighted!

    I was confused for a bit on what that satellite orbiting Moon is, and why the Moon is blue, before realizing how it works, so was a mini puzzle for me too :)

    • vunderba 7 hours ago

      Haha - I appreciate your generous interpretation of my rather poor UI/UX as a puzzle! That's going to be my excuse in the future.

senko 4 days ago

An experiment inspired by the "Gonon: building a clock with no numeral" recently shared on HN (https://news.ycombinator.com/item?id=47532338)

The end result is: https://senko.net/clock/

Tech trivia: implemented as a reusable web component, powered by CSS animation, JS is just used to set up the initial state.

  • orly01 18 hours ago

    Nice experiment.

    Some feedback:

    I could not catch that the circle meant the sun. I did catch seconds, minutes, day, month, and I deduced TK meant Tokyo and SF meant San Francisco and could not deduce the city ZG represents. So, I don' think you can actually be free of baggage. Some minor things bugging me: - The lines at the bottom don't align! Really annoying, I spent a lot of time trying to understand it meant something or not. - Dark mode shows only 1 city. I could not understand what TK meant until I saw SF.

    Some advice:

    - Change the pictoric representation of the sun for something a bit more "sunny". Some flames feel like something not cultural (since there literally is solar activity at the surface), so aliens could get that it is the sun. - In the intro you say there is a puzzle, but you don't say what the puzzle is. I thought it might be some interactive thing. It was only until I gave up I start reading the explanation and got that the puzzle is to understand how the clock works, given it has no cultural baggage. Then I went back to try to understand it. Maybe add a little explanation somewhere saying the intention.

    • senko 13 hours ago

      Thanks for the kind words and the feedback & suggestions! Yeah I don't think you can get completely free of the cultural assumptions.

voidUpdate 9 hours ago

Couldn't you just stack some continuous progress bars? The article is allowing use of hours, minutes and seconds, so just have a 1 minute bar, a 1 hour bar, a 1 day bar, etc. No numbers needed, and if you differentiate between "past" and "future", say by filling in the progress bar, direction is unambiguous

sixeyes 10 hours ago

The obvious thing for me is to have the orbit around sun, second, etc - all the things you decided to run clockwise - as fixed points, at the top, and let the tracks themselves run anti clockwise. Then everything would move the same way, same direction, and a "now" that's a the same spot always.

As it is now, "now" is at different places at the different circles.,.,.

kogold 11 hours ago

I think in the end the takeaway from both your and the Gonon - project is that timekeeping only makes sense if it is linked to commonly shared, clearly perceptible phenomena that occur regularly, because otherwise you either have no shared concept to communicate time or you always land at even more arbitrary conventions or social constructs.

  • senko 3 hours ago

    Yup!

    Turns out if you want to have a useful clock it needs to use the social/cultural constructs. This was not at all obvious to me before I started with the project, but now, especially after the discussions here and elsewhere, it looks unavoidable.

    Or to quote one of my favorite books: "Time is an illusion. Lunchtime, doubly so."

amarant 16 hours ago

This is awesome! I kinda wanna make one that adapts even more to your location: lengthen and shorten the light part of the circle according to your local sunrise and sunset times! Unfortunately that kinda sorta breaks the neatness of displaying multiple timezones, since it won't line up nicely with different latitudes.

  • senko 14 hours ago

    > lengthen and shorten the light part of the circle according to your local sunrise and sunset times!

    It's really not obvious, especially in light mode, but if you switch to dark mode you can see the day circle has a "bolder" and a "thinner" part. Bolder is the daylight hours, and its length and position corresponds to the daylight hours (in this instance, of Zagreb, since that's the primary timezone for the clock).

    You're correct in that it can't work if you have multiple lattitudes. I took the easy way out - I ignore that and just use the first one :X

    • amarant 12 hours ago

      Oh cool I didn't realize! I thought it was just half. Well, this time of year that's to be expected I guess

mungoman2 16 hours ago

Instead of anchoring the sun and thus noon at the top it would be interesting to have the sun move around the clock face as the year progresses. Noon then moves around as the year progresses. ”Up” could be said to point towards the center of the galaxy instead.

parasubvert 17 hours ago

Cool project.

I had thought that months aren't quite a human construct, they correspond roughly to lunar cycles. Weeks were a way to carve the month up into the four lunar phases per cycle.

Seconds, minutes, hours, etc. are, as you say, all sexagesimal math bias.

  • senko 14 hours ago

    You're right about months/weeks, I was imprecise in the post: calendar months (as we now use them) are a weird cultural construct (I mean who would divide a measurement unevenly and though that was a good idea?)

    Referencing actual phenomena would be more elegant - but wouldn't show correct date (see my comment here https://news.ycombinator.com/item?id=47686522)

  • gshiut 2 hours ago

    give me a virius

  • gshiut 2 hours ago

    anyone else ?

rimasu 13 hours ago

I've been working on something a little similar. It's also a clock with no numbers - but much nearer to a conventional analog clock.

Main changes are greater information density and minor modifications to the movement mechanism. Site has themed examples and a tutorial.

https://rimasu.github.io/orbitalclock/

  • senko 13 hours ago

    That's cool, thanks for sharing!

    Does the arc length signify anything? Wasn't able to figure it out, and as far as I can see, it isn't explained.

    • rimasu 13 hours ago

      Arc length is fixed at 5/12th of a circle.

      Deliberately not half so you can tell what is foreground/background irrespective of color scheme.

      I did explain design thinking in early versions of the tutorial, but it made it too long.

lmm 14 hours ago

> Note the day/night line is a bit tilted – it's not exactly perpendicular to the sun. This is because, as I write this, Croatia uses DST (Daylight Saving Time) – we intentionally shift time to have more daylight in the evening. While this shortens daylight in the morning, most people are asleep at that time so won't mind it.

Erm, WTF? DST does not change when dawn and sunset occur relative to solar noon. So what exactly is this showing?

  • senko 13 hours ago

    That's a good catch! The clock is inspired by the solar noon (uses sun as the anchor), but shows clock time, so it can be of practical use.

    Rereading the post, I notice I haven't made that distinction clear.

    • 0zer0 6 hours ago

      Really nice clock.

      One nitpick on the day/night line: This is strongly dependent on the latitude, i.e. it kind of collides with your notion of the world clock. Maybe you boldface the label for the city to which it is referring.

  • boxed 13 hours ago

    Came to write this. That sentence is clearly nonsense. Either you display the sun, or you don't. Don't display it in the wrong position!

altairprime 17 hours ago

Senko, note that dark mode hides the two secondary location labels.

This is a charming clock :)

  • senko 14 hours ago

    Whoops, thanks for reporting it. And for the kind words, happy you like it!

mauriciolange 11 hours ago

This article, being in turn inspired by another project, has a much more fun to read writing style. The "Gonon: building a clock with no numeral" is full of AI slop and that is very distracting.

so, kudos to the author at Senko.net!

  • senko 6 hours ago

    Thank you, appreciate the kind words!