Wikipedia talk:Manual of Style/Dates and numbers

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Manual of Style
WikiProject icon This page falls within the scope of WikiProject Manual of Style, a drive to identify and address contradictions and redundancies, improve language, and coordinate the pages that form the MoS guidelines.

What is the correct unit symbol for gigabaud?[edit]

The unit baud (symbol Bd) is frequently written with (decimal) prefixes kilo, mega and giga. The logical unit symbols for the resulting kilobaud, megabaud, gigabaud are kBd, MBd and GBd. I recently changed Gbaud to GBd to conform with this logic, and my edit was reverted. What is preferred, GBaud, Gbaud or GBd? Dondervogel 2 (talk) 14:43, 8 December 2014 (UTC)

Yeah, I'd also like to know that. :) — Dsimic (talk | contribs) 14:49, 8 December 2014 (UTC)
Hi. Glad you found your own way here. I will also flag the discussion at baud.
In the International System of Quantities "the baud may be combined with prefixes, for example kilobaud, symbol kBd, megabaud, symbol MBd", from which I deduce that the correct ISQ symbol is GBd. I see no good reason to depart from the ISQ myself. What do others think? Dondervogel 2 (talk) 14:57, 8 December 2014 (UTC)
Why should we treat this any differently than we do WP:Manual_of_Style/Dates_and_numbers#Quantities_of_bytes_and_bits (with all the attendant decimal/binary confusions we've somehow learned to live with)? Or is the question simply what the proper symbol is for baud? EEng (talk) 18:04, 8 December 2014 (UTC)
I think the question is only what the proper symbol for baud is. The fundamental reason the capacities of RAM and ROM are normally a power of 2 is that when an additional memory address line is added to a memory chip, it has generally been the practice to take full advantage of the additional possible addresses. This does not apply to data transmission, so there is no particular reason why the number of symbols in one second should be a power of 2, and hence, no reason to assign binary meanings to prefixes when working with data transmission rates. Jc3s5h (talk) 19:05, 8 December 2014 (UTC)
Whether there's a sensible underlying reason or not, lots of computing stuff is measured in powers of 2 for no reason other than that lots of computing stuff is measured in powers of 2. (Repeat) EEng (talk) 22:11, 8 December 2014 (UTC)
But data transmission rates are not among that "stuff". A "56K" modem is running at 56,000 bits/sec, and a "100 Mbit" Ethernet link at 1,000,000 bits/sec. Even the clock rates on RAM are quoted using decimal prefixes, even though RAM capacity is quoted in powers of 1024. Jeh (talk) 12:18, 9 December 2014 (UTC)
A 56K modem is actually 57,600 bits/sec. Data transmission rates were typically power-of-2 multiples of 75 baud: 75, 150, 300, 1200, 2400, 4800, 9600 etc. (I never heard of a 600 baud modem, but that doesn't mean they didn't exist) and later power-of-2 multiples of (I think) 3600 baud: 3600, 7200, 14,400 etc.; and 57,600, being 14,400 * 4 is in that sequence. The power-of-2 multiples are because of clock rates: given a timing crystal of a certain characteristic frequency, it's a simple matter to add divide-by-2 or multiply-by-2 circuits to alter that frequency; but the resultant frequency will be a power-of-2 multiple of the original. To raise a frequency by ten times is more difficult than to raise it 16 times.
As for RAM sizes, they're always exact powers of 2 - it's not so much "to take full advantage of the additional possible addresses" but because it's easier to make a device where every possible address is accessible than to make one that is smaller, yet larger than half the size. 1,000,000 byte devices are never made, but if they were, the easiest way would be to make a 1,048,576 byte device and add some circuitry to mask off (and therefore ignore) addresses greater than 0x0F423F - which would be a waste of those 48,576 locations in the range 0x0F4240-0x0FFFFF. --Redrose64 (talk) 13:23, 9 December 2014 (UTC)
Aren't modem-to-modem data transmission rates actually given in bits per second? Thus, for example, it would be 14400, 28800, 33600 or 56000, but bit/s, not baud; in bauds, those would be 2400, 3000 or 3200 baud, except for the 56k modems that stretch the limits of phone lines and have different uplink and downlink speeds. Only the early modems (up to 2400 baud or so) did have equal numeric values for their symbol (baud) and bit rates (bit/s), meaning that for them one transferred symbol equaled one transferred bit; later, that changed with the advancements in modem technology so increasingly more bits were encoded and transferred per symbol. Here's a reference for more information. — Dsimic (talk | contribs) 13:59, 9 December 2014 (UTC)
Dsimic is correct, the common usage is misleading. ADSL signalling uses even more indirect representation, with multiple signalling channels, each having its own symbol rate, symbol encoding constellation, noise margin, error correction, etc. The data rates from all the channels are aggregated, so it I suite wrong to refer to a baud rate for what is really the aggregate data rate.LeadSongDog come howl! 14:27, 9 December 2014 (UTC)
Yes, I meant to mention the baud - bit/s confusion. But look, this conversation belongs over on the talk page of the article in question, among editors familiar with the sources. As far as we know that's the only place this issue has come up, and it should be resolved there. EEng (talk) 17:07, 9 December 2014 (UTC)
The original issue question was whether one should use Gbaud, on the grounds that it is frequently used (familiar) or GBd, which is prescribed by international standards. I don't think this is the same question as the correct symbol for baud, but we can add that in as well, so here goes:
  1. What is the correct symbol for one baud? Bd
  2. What is the correct symbol for 1000 baud? kBd
  3. What is the correct symbol for 10002 baud? MBd
  4. What is the correct symbol for 10003 baud? GBd
The ISQ's answer to these questions is given after the question mark, but in its wisdom mosnum proudly chooses to flout ISQ. What does mosnum say? Dondervogel 2 (talk) 20:32, 8 December 2014 (UTC)
I'll give my standard response: this should be worked out on the talk page of the article in question. If this becomes a recurring issue on several articles, the maybe MOSNUM should have a provision added. Having said that, I think it's probably a bad idea to mix abbreviated prefixes (K, M, G) with full unit names (baud). So I think megabaud or MBd, but not megaBd, Mbaud. What do sources use? EEng (talk) 22:11, 8 December 2014 (UTC)
Having all this in mind, I'd say that using megabaud and gigabaud would be the best option, as (so far) I've never seen MBd or GBd in a text. — Dsimic (talk | contribs) 22:35, 8 December 2014 (UTC)
Yes, gigabaud is fine in text, but sometimes a symbol is appropriate. I'm not sure that following any old source is wise. After all, Gbps is more common than Gbit/s for "gigabit per second", but I prefer Gbit/s because that is the international standard symbol. By the same token Gbd is the international standard symbol for gigabaud. As for discussing this issue on every talk page, that would never lead to uniformity, which surely is mosnum's primary goal? Dondervogel 2 (talk) 22:54, 8 December 2014 (UTC)
First, I'm not suggesting following any old source -- I'm suggesting that inquiring into sources' usage is a good place to start the discussion. Second, I'm not suggesting discussing this issue on every talk page (of every article in which the question comes up, I guess you mean) but rather ... OK, here's my standard spiel:
It is an axiom of mine that something belongs in MOS only if (as a necessary, but not sufficient test) either:
  • 1. There is an apparent a priori need for project-wide consistency (e.g. "professional look" issues such as consistent typography, layout, etc. -- things which, if inconsistent, would be noticeably annoying, or confusing, to many readers); OR
  • 2. Editor time has, and continues to be, spent litigating the same issue over and over on numerous articles, either
  • (a) with generally the same result (so we might as well just memorialize that result, and save all the future arguing), or
  • (b) with different results in different cases, but with reason to believe the differences are arbitrary, and not worth all the arguing -- a final decision on one arbitrary choice, though an intrusion on the general principle that decisions on each article should be made on the Talk page of that article, is worth making in light of the large amount of editor time saved.
EEng (talk) 04:25, 9 December 2014 (UTC)
Just as a note, I also prefer Gbit/s over Gbps. At the same time, it might be good to have a baud-related guideline, though I'm not sure how often it pops up? — Dsimic (talk | contribs) 23:17, 8 December 2014 (UTC)
A committee somewhere may have an answer for this, but in an encyclopedia the question is irrelevant. Just write "gigabaud" so there is a chance some readers will know what it is (although the 109 vs. 230 issue would need to be resolved from context). Johnuniq (talk) 23:34, 8 December 2014 (UTC)
I'm sensing there is no consensus on choice of symbol. On the other hand there seems general agreement on avoiding this issue by spelling out the term gigabaud in full. I will try that and see how it goes. Dondervogel 2 (talk) 08:58, 9 December 2014 (UTC)
Sheesh. I'm surprised at how much debate there is about this. But then, I'm surprised at the number of editors that feel that (even marginally) "common usage" is the overriding factor in the MOS. Makes me wonder what the purpose of the MOS is, other than as stabilizing tool in arguments. —Quondum 16:05, 9 December 2014 (UTC)
The purpose of MOS is (indeed) primarily to be a stabilizing tool to avoid arguments over arbitrary, unimportant stuff. See my text in italics above for a more complete formulation, at least as I like to express it. EEng (talk) 17:07, 9 December 2014 (UTC)
In your opinion. IMO, it is to set and maintain a relatively uniform style, which is not at all the same thing as reducing arguments between editors of differing tastes. And I note you do not address the point of "common usage" at all. This is not entirely at odds with your italicized text: I agree that the MOS should not be unduly prescriptive and detailed, but where a diversity of conflicting styles would otherwise proliferate, even without conflict, I feel that it would be valuable. Otherwise we could simply use the simple rule: first major author of an article dictates the style for that article. There are lots of ways of resolving editorial conflict if that is the primary objective. —Quondum 17:47, 9 December 2014 (UTC)
Um, what I said is that, IMO, MOS' purpose is primarily to be a stabilizing tool. My point 1. (in my italics text) exactly speaks to your "relatively uniform style" concern. But for the moment this seems to have come up on one article only, so uniformity isn't an issue as yet, and if it is, we don't know what additional articles are involved. For the nth time, let editors hash this out on this one article's talk page. If the result is peaceful, they might make an effort to export it to other related articles where the agreed-upong notation might be used. And that would be the end of it -- no need to bloat MOS until/unless it really seems like there's a problem to be solved across multiple articles.
We do use first-major-contributor as the tiebreaker in a few cases (WP:ENGVAR) but in general that's not going to give us good results. EEng (talk) 19:30, 9 December 2014 (UTC)
Okay, I probably misinterpreted you. I guess this might be regarded to fall between the cracks (of the MOS as it stands) as a non-SI unit; I tend to assume the SI naming rules would still apply, though, but I'm not sure whether the MOS deals with this. —Quondum 20:51, 9 December 2014 (UTC)
For the record, I disagree that this is best dealt with on individual article talk pages because there are too many of them. If there is a suitable Wikipedia Project talk page, that might be a good alternative for a centralised discussion. Is there one? A separate issue is the (in my view) undue weight placed on common use. In my opinion the discussion could be simplified if there were a general advice to follow the ISQ unless there is a good reason not to. That way it is only the exceptions to the ISQ that would require discussion, thus saving the presently disproportionate amount of editors' time on discussions like this one. Dondervogel 2 (talk) 21:23, 9 December 2014 (UTC)
What "too many of them" -- again, so far only one article has been mentioned on which this has arisen. When/if it starts coming up in other articles, then we might consider getting MOS involved. EEng (talk) 01:28, 10 December 2014 (UTC)
With some assistance from Dsimic, I have just edited several dozen to make them consistent. What are the chances of these articles remaining consistent without a guideline, and what are the chances of avoiding new inconsistencies without a guideline? Dondervogel 2 (talk) 08:42, 10 December 2014 (UTC)
With or without a guideline in MOS on the subject, the probability of all these articles remaining consistent is 0 unless some editor takes an interest in keeping them consistent, because MOS guidelines don't emit magic enforcement rays which exert their power on articles. There may or may not be trouble in maintaining consistency (and, BTW, it's not even certain consistency is desirable on this -- I dunno, maybe different articles have different needs) but if so, then interested editors from the various involved articles can bring their ideas here for an informed discussion, including examples from sources, standards bodies, etc. Right now what we have here is an abstract discussion, solving a hypothetical problem, engaged in by editors who mostly have no more than a passing interest in the subject. That's not a good way to develop guidelines. If this actually becomes a problem, bring it back here. EEng (talk) 11:20, 10 December 2014 (UTC)
Just as a note, I have more than just "drive-by editing" interests here. :) — Dsimic (talk | contribs) 20:16, 10 December 2014 (UTC)
I referred "editors who mostly have no more than a passing interest" -- if that's not you, great. But most here are just the usual MOSNUM regulars. That's not a put-down, just that those actually engaged on the article in question are in the best position to bring to bear everything that may be useful in resolving the question. EEng (talk) 22:23, 10 December 2014 (UTC)
Thanks for the clarification, it makes sense. — Dsimic (talk | contribs) 22:37, 10 December 2014 (UTC)
Well, usability should be the paramount. I'd bet that very few readers would figure out what GBd means (could be seen even as a misprint of GB), while the meaning of gigabaud should be easily understood. — Dsimic (talk | contribs) 22:02, 9 December 2014 (UTC)
The same could be said of almost any other SI or ISQ unit you care to mention. Linking GBd is sufficient for usability. I'm 100% with Dondervogel, including that we should extend the MOS statement from SI to ISQ. —Quondum 22:55, 9 December 2014 (UTC)
I still think that using GBd, linked or not, would favor form over function. — Dsimic (talk | contribs) 23:03, 9 December 2014 (UTC)

Do we add a space in >_10 °C?[edit]

Do we write >10 °C or > 10 °C? I also want to use it with conversions: >_10 °C (50 °F; 283 K). Inequality (mathematics) did not help me out (at all). -DePiep (talk) 10:46, 19 December 2014 (UTC)

FWIW, I'd say that it would be the best to write it as "higher than 10 °C" or "over 10 °C". — Dsimic (talk | contribs) 11:21, 19 December 2014 (UTC)
> 10 °C is better. Dondervogel 2 (talk) 12:07, 19 December 2014 (UTC)
It's straight from the ISO ruling—a pretty good one, in my view, and from long ago. Tony (talk) 12:26, 19 December 2014 (UTC)
I'm working at an infobox, so verbose is not preferred. -DePiep (talk) 12:41, 19 December 2014 (UTC)
Then it should be spaced, though "over" isn't that much longer either. — Dsimic (talk | contribs) 12:46, 19 December 2014 (UTC)
In science the symbol is quite common (chemistry). Other signs used are < and ~. I am not aiming to prescribe word/symbol usage in this. -DePiep (talk) 13:08, 19 December 2014 (UTC)
Yeah, I know what these symbols are for. :) Didn't want to "derail" your question, sorry. — Dsimic (talk | contribs) 13:12, 19 December 2014 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── ISO said it, so a space there will be. -DePiep (talk) 13:08, 19 December 2014 (UTC)

ISO is a private not-for-profit organization known for charging exorbitant fees for there standards, so it is not an ideal organization to rely on when deciding style matters in Wikipedia. They have no authority to enforce any of their standards, and many of their standards languish because no one pays any attention to them. We use a space because this guideline says so; in the guideline, search for "40 °". Jc3s5h (talk) 15:30, 19 December 2014 (UTC)
Yes, ISO keeps their standards secret, in Swiss vaults. A great way to promote it. The guideline you point to (as I fou nd it) is about the space in "40_°" (don't) vs. "40_°C" (do), which is not what I am looking for. It's about ">_10 °C?" or ">10 °C". Since we have no guideline on that, I'll go for second best: a sensible reply here. (If it were more strong in one way or another, that would be clear too). -DePiep (talk) 16:03, 19 December 2014 (UTC)

────────────────────────────────────────────────────────────────────────────────────────────────────` Sorry to be dense, but what does the _ mean as used above? EEng (talk) 17:36, 19 December 2014 (UTC)

A "visible" space character. — Dsimic (talk | contribs) 17:39, 19 December 2014 (UTC)
It is the position of the space/nospace we are talking about. -DePiep (talk) 17:54, 19 December 2014 (UTC)
Thanks, I get it. BTW, if there's a space after the < it should probably be an nbsp i.e. <&nbsp;10&nbsp;&deg;C EEng (talk) 18:47, 19 December 2014 (UTC)

Notice of a related discussion[edit]

For those of you who still interested in this subject; there is a discussion in progress at Wikipedia talk:WikiProject Football#Why meters first for English player height? in which systematic unit flipping, source-based units, source selection based on preferred units, reasons to ignore advice in WP:UNITS & co. are being discussed in relation to the unit order to use for the players in English football (soccer). Baaarny (talk) 07:36, 22 December 2014 (UTC)

Lasciate ogni speranza, voi ch'entrate. EEng (talk) 07:47, 22 December 2014 (UTC)

Suggested addition[edit]

In the table of date formats that are not recommended, I suggest adding all-numeric examples with the year abbreviated to 2 digits. In the world outside Wikipedia, one still sees forms like 10/11/12 that are maximally ambiguous. I believe even ISO 8601 allows 07-04-15 (meaning April 15, '07) if the leading digits can be inferred, but that's not a proper encyclopedic style. -- (talk) 06:09, 29 December 2014 (UTC)

I thought about how to add this warning, but honestly it makes the table just that much more bloated for what I think is little benefit. I really don't care what ISO blah-blah says -- has there actually been a real problem, in a real article, where editors argued over this? It's perfectly obvious from nearby examples that "all-digit" years are required, and editors are not morons who need everyone absolutely spelled out. EEng (talk) 11:06, 29 December 2014 (UTC)