Talk:12-hour clock

Table in noon and midnight section has incorrect entries

The NIST line is wrong. 12:01 am is 12:01 am. How else can you distinguish 12:01 am from 12:00 am? If it were 12:01 am and someone asked you the time, you would say 12:01 am. But 12:01 am is midnight according to the table. So when someone says it's 12:01 am, is it 12:01 am or midnight? This is ridiculous.

You use 12:01 am and 11:59 pm only when down-to-the-minute accuracy is not needed. While NIST doesn't explicitly say this, it is strongly implied (use in contracts is mentioned) and the only sensible interpretation. The table should reflect this.

It's nice to see the U.S. gov't finally got it right! (^_^)

Also, how about an entry for computers? It's not just digital watches. In fact, it's not just computers. It's digital displays in general.

Betaneptune (talk) 19:40, 19 November 2019 (UTC)

Are there any objections to changing "digital watches" to "digital displays" or "clocks with digital displays" or "digital clocks"? The first one is really too specific. Betaneptune (talk) 20:54, 20 October 2020 (UTC)

I don't mind changing "digital watches" to "clocks with digital displays" but you should start a new section rather than sticking this in the middle of a stale thread. Jc3s5h (talk) 22:36, 20 October 2020 (UTC)

Yes, it could be clearer. The intention, of course, is that 11:59 p.m. lasts up to 11:59.99; there is no midnight (the contract time changes at the instant of midnight from 11:59 to 12:01 a.m.); and 12:01 starts at 12:00.01

How can we express this more clearly? This is for contexts where the date is more important than the time to the exact second. Dbfirs 21:08, 19 November 2019 (UTC)

My only concern is that the table possibly gives the mistaken impression that 12:01 AM and 11:59 PM are midnight and noon, respectively. Perhaps I'm going too much by how the article was years ago, the last time I took an active part in it, when it was a disaster. My impression here is that 12:01 AM and 11:59 PM are being used as close enough to midnight and noon for most purposes, but still different, so as to make clear what day is meant. But the article and/or table isn't making that clear. So why not just state that? That is clearly the intent of the NIST bit. --Betaneptune (talk) 21:49, 19 November 2019 (UTC)

Noon isn't involved here. The explanation is in the adjoining text. Dbfirs 21:59, 19 November 2019 (UTC)

Right. Sorry. Strike noon. Should be beginning of day instead. Thanks. (^_^) Betaneptune (talk) 22:43, 19 November 2019 (UTC)

I support adding a row for computers, specifically spreadsheet date formats. - Msramming (talk) 18:02, 5 September 2024 (UTC)

Computers hold time in seconds (or even microseconds) since a base moment last century (see Unix time). How that is presented in application and locale specific (for example, the US uses MM DD, YYYY; most of Europe uses DDMMYY, Japan uses YYYYMMDD). On a spreadsheet, you can equally choose how time is displayed. But it wouldn't be "computer time", it would be an application-specific presentation of time, so the line (if agreed we should have it) would be Microsoft Excel displays: ... . 𝕁𝕄𝔽 (talk) 19:15, 5 September 2024 (UTC)

12 AM is noon and always has been

12 AM has never been midnight, and it doesn't logically make sense for it to be so why are 12 AM and 12 PM swapped 2601:242:4100:F200:418B:F12F:5C3:F4C0 (talk) 00:38, 24 February 2021 (UTC)

Every Wikipedia editor is unreliable. All contentious statements made in Wikipedia must obey the verifiability policy. A reliable source, the 2008 edition of the US Government Printing Office Style Manual is cited in the article and says that 12 AM is midnight of the day that is beginning. The 2016 edition says the same thing on page 275 (which is page 289 of the PDF). Since there is a significant source that says 12 a.m. is midnight, your assertion "12 AM has never been midnight" is just wrong. Jc3s5h (talk) 00:50, 24 February 2021 (UTC)

You are wrong about 12.00am and 12.00pm. The thing is that 12.00am comes directly after 11.59am and is am all the way through until it becomes 1.00pm - 12.00am - 12 midday - until the afternoon - 1 in the afternoon. Likewise, 12.00pm is 12 o'clock midnight and the time only becomes am (morning) when it becomes 1 in the morning - 1.00am. People have always got this wrong and I am surprised that Wikipedia is also making the same mistake.

Many thanks, Dave Rattle — Preceding unsigned comment added by 102.65.62.161 (talk) 07:37, 28 May 2021 (UTC)

Please have a look at any digital clock at any time between noon and 1 pm. What does the am/pm marker say? Case closed. −Woodstone (talk) 12:38, 31 May 2021 (UTC)

Dave Rattle, 1) Don't put your contact information into discussions here. 2) If it is indeed am until 1:00pm, what do we make of 12:01 after noon? How about 12:00:01? 12:00:00.1? 12:00:00.00000....[insert any finite number of zeros you like]....0001? All of those times come AFTER noon, right? Uporządnicki (talk) 16:20, 8 September 2021 (UTC)

The day starts at 12:00 am. Later we get 12:01 am, 12:59 am, 1:00 am, . . ., 11:59 am, 12:00 pm, 1:00 pm, . . ., 11:59 pm, 12:00 am. The most logical, the most-often used, and in fact, the de facto convention is that 12am is midnight at the beginning of the day and 12pm is noon. Every digital clock I've ever seen, including embedded systems, follows this de facto convention. If you need to set a clock or alarm to 12:00 on a 12-hour digital clock or embedded system, you better know and use the de facto convention, or you'll get it wrong! And it has never been the case that we switch from am to pm or vice versa at 1:00.

The de facto convention is the most sensible, the most logical, follows the usual rollover rules with AM switching to PM and vice versa at 12:00, convention for the 12-hour clock. It also follows the KISS credo. Anything else is more complicated. And insisting that 12am and 12pm are illegitimate symbols for midnight and noon shows a slavish adherence to the definitions of am and pm that were made approx. 400 years ago. All other words are allowed to evolve. But for some reason, some people (fuddy-duddies), insist on using the original outdated, archaic definitions. It's well past, uh, time, to move on! Why make it more complicated than it has to be? Betaneptune (talk) 10:12, 2 March 2024 (UTC)

If you want to make people do what you think is logical, get yourself appointed Secretary of Commerce or Secretary of Defence of the United States. Then you'll be in charge of one of the two agencies in the US responsible for time dissemination, and a few hundred million people might care what you think. Until then you're an anonymous Wikipedia editor, with the degree of recognition that goes with that position. Jc3s5h (talk) 16:40, 2 March 2024 (UTC)

The day definitely does NOT begin at 12am... the day is divided by noon at the meridian... so am is ante meridiem (the time before midday) and post meridiem (the time after midday). Midnight is definitely after midday so is 12pm. 12am is noon. Using the 24hr clock, 0001 is 12:01am Marlarkey (talk) 23:47, 9 October 2024 (UTC)

Well, excuse me for putting an entry in the talk space! I didn't know it was your personal playground.

What I think is logical? Tell me how what I said is not logical. Go ahead. I know, it's those 400-year-old definitions of am and pm. Well, most clocks are digital today, and when the clock says 12:00 pm, you need to know whether it's the middle of the day or middle of the night. And, gee, what do you am/pm sticklers do when you have to set an alarm? I guess you do the 11:59 and 12:01 trick. Fine.

Ah, again with the stupid ante meridiem and post meridiem stuff. These words are from hundreds of years ago. Those people didn't even have electricity, much less vacuum tubes (aka radio tubes), much less modern electronics (they contain transistors). Apparently, all words are allowed to evolve except these two.

And you're not an anonymous Wikipedia editor, Mr./Ms. Malarkey? I guess your posts are also malarkey.

Why the hostility? All I did was explain what already is the standard for digital clocks (including the times displayed on a computer). It is people like you why I have stopped donating to Wikipedia.

This whole article is ridiculous. Are we next going to have an article describing the intricacies of addition and subtraction, explaining carrying and borrowing? There shouldn't even be an article like this.

And Mr./Ms. Marlarkey, when you get an email timestamped 12 am or 12 pm, are you genuinely confused? I suggest getting some training.

Fine. Stick with your archaic definitions. Have a snit about am and pm. And the rest of the world will pass you by with our digital electronics. Betaneptune (talk) 03:02, 11 December 2024 (UTC)

Correct... pm = post meridiem in latin and 12 midnight is after midday so is definitely 12pm. If midnight is 12pm then noon must be 12am. Marlarkey (talk) 23:44, 9 October 2024 (UTC)

Another of looking at this is to say which day does midnight belong to ? If I say something has to be done by midnight on Friday then it be done anything up to midnight on Friday ie that midnight is part of Friday. To suggest that midnight is part of Saturday makes it nonsensical to say "midnight on Saturday" when you mean by midnight at the end of Friday. So 12 midnight is part of the day that it is at the end of. And it is after midday hence post meridiem so midnight is 12pm. And therefore noon is 12am. Simple. Marlarkey (talk) 23:59, 9 October 2024 (UTC)

Please expect WP:V, WP:NOR, and WP:IRS to be strictly enforced in this article. Did you read the chart which gives different meanings adopted by various reliable sources? Since reliable sources, there is no single correct answer. Jc3s5h (talk) 01:20, 10 October 2024 (UTC)

Since reliable sources, there is no single correct answer. Presumably "Since reliable sources differ, ...". Guy Harris (talk) 21:25, 5 November 2024 (UTC)

To add reference of a day start/end as per hindu calendar

It is notable to add the references of how the Hindu calendar defines the start of a day and end of day.

Extract from the page (Section: Sandhyakala in https://en.m.wikipedia.org/wiki/Sandhyavandanam) As per Hindu traditional calendar, the day starts with sunrise (i.e. from midnight of previous night until sunrise is considered part of previous day).

Thaejas (talk) 00:32, 11 September 2022 (UTC)

This article is about the 12-hour clock. How would you add information about the traditional start of a day in the Hindu traditional calendar to this article? Also, Wikipedia is not a reliable source so you would have to find reliable sources for whatever you want to add. Jc3s5h (talk) 17:27, 12 September 2022 (UTC)

How did the Romans (and Egyptians) divide the night?

The article claims that the night is divided into four "watches", but the reference for the "day divided into 12 variable-length hours, night divided into four watches" claim appears to say, instead, that the night is also divided into 12 variable-length hours. Which is it? Guy Harris (talk) 07:24, 5 May 2025 (UTC)

From the diagrams in Roman timekeeping § Subdivision of the day and night, the night appears to be divided into 12 units (presumably hours), with each group of three hours being part of a larger subdivision. The caption mentions vigiliae, which I suspect is the plural of "watch"; presumably those four groups of three hours are the "watches". Guy Harris (talk) 07:33, 5 May 2025 (UTC)

The only source cited ( Smithsonian Magazine) simply says

Ancient Romans didn’t measure time in our 60-minute hours; instead, they divided daylight and darkness into 12 increments each, a system they adopted from the Egyptians. In Rome, that meant an hour was about 45 minutes in winter and 75 in summer. Hours would have governed meetings, courts and dinners, but not in the carefully structured way so many of us experience today.

So unless another citation is provided, the "four watches" text must be deleted. 𝕁𝕄𝔽 (talk) 07:47, 5 May 2025 (UTC)

Found it. "The best-known instance of temporal organization inside the Roman army is the night watch (vigiliae), the night was divided into four equal parts of three hours each, with the help of the Klepsydra." I'm not sure to what extent any person or organization other than the army (or other organization that needs to keep, well, watch during the night) cared about vigiliae or hora during the night, but I'll just note the grouping of night hours into watches. Guy Harris (talk) 07:51, 5 May 2025 (UTC)

Done. Guy Harris (talk) 08:02, 5 May 2025 (UTC)

The existence of 24:00 or 2400

In a recent edit summary, JMF stated "24:00 is never used in 24-hour notation. Military uses 23:59 to prevent misunderstandings."

On 15 May 2025 the United States Department of the Air Force published a memorandum, DAFMAN11-401 Aviation Management. This memorandum uses 2400 on page 65. Jc3s5h (talk) 18:07, 26 August 2025 (UTC)

How standards have slipped! () 𝕁𝕄𝔽 (talk) 18:42, 26 August 2025 (UTC)

You sometimes need the actual midnight rather than minute before or after. Even ACP 121 which is used as a ref in this article used "2400" to refer to the midnight exactly. It said not to use 0000 and only to avoid 2400 "unless it is necessary to indicate this particular instant of time". – MwGamera (talk) 20:30, 26 August 2025 (UTC)

Hi, I'm an anonymous user who made an edit to the page in question. I was the one who added 24:00 (twenty-four o'clock) to the table, but I made a mistake because 24:00 (twenty-four o'clock) doesn't exist. In the 24-hour format, midnight is represented by 00:00 (zero o'clock). User JMF is correct about removing 24:00 (twenty-four o'clock) from the table.2804:214:8118:3D3A:B476:3D06:EF29:967 (talk) 00:08, 28 August 2025 (UTC) — Preceding unsigned comment added by 284:8118:3D3A:DDF7:71B9:95C7:965C (talk • contribs)

I don't have opinion regarding its inclusion in the table; inclusion seems likely to be more contentious than omission and it hardly matters in the article about 12-hour clock. But it really depends who you ask and according to which standard. Just like with am/pm at noon and midnight. Saying that 24:00 "doesn't exist" or that it's "never used" without any further qualification is simply false. Either is allowed in ISO 8601, for example. (@Jc3s5h: was there any edition that did not allow it?) Some profiles like that of W3C or RFC 3339 indeed limited it to allow only 00:00. But in military contexts, conversely, 2400 (twenty-four hundred) seems to be more common, at least in the past. Look at the sources under 24-hour clock § Midnight 00:00 and 24:00. If you follow, for example, ACP 121 or the older version of SECNAV M-5216.5 from before 2015, then it's 0000 that "doesn't exist". It's not a slip. – MwGamera (talk) 14:17, 28 August 2025 (UTC)

I conceded the point given that Jc3s5h cited the RS. But I am surprised. If an attack is to begin "at midnight on Sunday" or worse still "24:00 on Sunday" or "00:00 on Sunday", is that Saturday night/Sunday morning or Sunday night/Monday morning. 23:59 Sunday is unambiguous. Though I suspect a wise general would avoid it completely and choose for 04:00. So my "it doesn't exist" really means "I've never seen it" which is (a) OR and (b) evidence that anybody sensible steers well clear to avoid the confusion. 𝕁𝕄𝔽 (talk) 15:07, 28 August 2025 (UTC)

A more common example would be a change in a regulation that goes into effect on a certain date. Although the original regulation probably just specifies a date, the local commander may want to interpret it for his/her subordinates by including a time of day. The time of day might be 0000 local time, or perhaps 1900 local time the day before what is stated in the regulation if the regulation is controlled by UTC. Technically, the local commander might not have the authority to change the effective date & time by a minute. Jc3s5h (talk) 17:04, 28 August 2025 (UTC)

The difference between 24:00 and 00:00 is only relevant if you also include the date. Like
Every night at 00:00(or 24:00) the batchjob starts.
On the 21st of February at 00:00 (implying this as 21st starts)
2025 ends on 31st december 24:00

IM(notso)HO — Preceding unsigned comment added by 37.96.98.88 (talk) 10:23, 18 September 2025 (UTC)

Jc3s5h and JMF: I think it's better to make this clearer and not add 24:00 to the table, as readers might get confused and think the same day has two midnights. The day actually starts at 00:00 (12:00 am) and ends at 23:59 (11:59 pm). If 24:00 exists, how would it be converted to 12-hour format? 12:00 am is synonymous with 00:00. One minute after 23:59, it is 00:00 the next day. It makes no sense to say that 24:00 on December 31, 2025 is the same as 00:00 on January 1, 2026, since the 24:00 time slot does not exist. Just as there are no times where the minute is the number 60. I know that the 24-hour format works by adding 12 more in the hours from 1:00 pm to 11:59 pm, and in the case of midnight, it is different, as midnight is the beginning of the day, so the correct way is to restart the count, subtracting 12 by 12, which results in 0, that is, 00:00. Midnight (12:00 am or 00:00) is the first hour of the day, not the last. The last hour of the day is 23:00 (11:00 pm). If midnight were the last hour of the day, then it would make sense for it to be 24:00, if the day started at 01:00 (1:00 am). I know that on analog clocks in 24-hour format, it is possible to see the number 24 representing midnight and even so, this is incorrect, as the correct number is 0 (or 00) representing midnight on these analog clocks. As I said before, I was the one who put 24:00 in the table and I recognized my mistake. 24:00 does not exist. Just as days in number (1 to 31), days of the week (Sunday to Saturday), months (January to December) and years are the same for the entire planet Earth, so the schedules follow this same logic. So, rethink the inclusion of 24:00 in the table. The difference in time is that there are two formats, 12-hour (AM/PM) and 24-hour. If 12:00 am and 00:00 are the same time, what would 24:00 be in 12-hour format? 2804:5758:0:4AA4:EC63:C1F8:9A33:B726 (talk) 14:08, 18 September 2025 (UTC)

Jc3s5h and JMF: It's the same thing as me saying that 00:00 on January 1, 2026 is the same thing as 23:60 on December 31, 2025. Times with minutes represented by the number 60 do not exist, as I said before. 2804:5758:0:4AA4:EC63:C1F8:9A33:B726 (talk) 14:11, 18 September 2025 (UTC)

"...readers might get confused and think the same day has two midnights." Each day has two midnights, one at the beginning and one at the end. Only reliable sources matter, not the opinions of Wikipedia editors. Nobody believes declarations by Wikipedia editors that something doesn't exist. Jc3s5h (talk) 14:38, 18 September 2025 (UTC)

You are of course strictly correct. Nevertheless, Wikipedia reports what is actually used, not should be used. See also WP:righting great wrongs. 𝕁𝕄𝔽 (talk) 14:56, 18 September 2025 (UTC)