In many parts of the world it will e.g. turn from 23:59 Friday 17th to 00:00 Saturday 18th in the middle of lunch.

This means a simple term like "lunchtime on Saturday" is suddenly ambiguous/useless, since it could mean around 00:30 Saturday (shortly after local solar noon) or it could equally mean around 23:30 Saturday (shortly before local solar noon the following solar day).

We don't have that problem with "Midnight on Saturday", so we'll solve it for "lunchtime on Saturday"

We do have the midnight problem. Every time I say something like, "Midnight on Saturday" the response is "Friday night or Sunday morning?"

Yes, midnight technically "belongs" to the day that follows, but it's often used informally to mean the end of the day.

Also, "midnight" has a defined time. It's 12:00 AM. "Lunchtime" does not have a defined time. It's sometimes 11:00 AM, or sometimes 12:30PM. It's a fuzzy time of day.

And the problem isn't just with lunch on a "split" day. What about phrases like, "After work on Friday" (when Friday begins at solar 17:00)? Is that gonna be a different day for Suzie who ends her shift at 23:00 UTC vs. Tommy who clocks out at 01:00 UTC?

What would quickly happen is that people would still use the Sun's position to demarcate the days, and we're back to days starting at different times of the clock in different places. And back to confusion when scheduling across large distances.

Look, the Earth is round and the sun rises at different times for everyone. Having discrete time-zones is probably the most elegant way of dealing with that problem. Any attempt to enforce UTC across the world just moves all the math to the shadows. We're still gonna have to do it, though. At least let's standardize it.

Yes we do..... if someone said that I'd ask if they meant "is that Friday evening or Saturday evening?"

(using "evening" because I assume more people stay up past midnight than wake up before midnight)

Thanks! That's the first decent argument I've heard against this.

Then again, people would develop a convention within days. A month, tops.

For example having local time vs official time. This is what remote cities in western China are using. They've to use same time as east coast, but it makes no sense at all for them. So they plan everything around unofficial local time.

That's an option, but I was thinking we would probably agree to call the days we've always called Thursday "Thursday" locally.

How do you know which was originally Thursday the earlier or later part? Or do you have to calculate it based off old UTC offset? What if you move across the globe? What if you talk to somebody online and they say they're doing smth on Thursday? How do you know which Thursday is theirs?

> Then again, people would develop a convention within days. A month, tops.

That's exactly my point -- the convention they adopt will most likely be to abandon UTC in favor of some local time. At which point, we end up with an ad hoc reinvention of timezones.

Let's say the world decides to follow UTC. In Hawaii, Midnight UTC is 2pm in the afternoon in our current system. So in the early "solar afternoon" it will switch from Monday to Tuesday.

So if you're working late on Tuesday you'd have to clarify or use some other phrase.