Wikipedia:Reference desk/Archives/Miscellaneous/2017 October 31

I heard that the Lamborghini engine (at least in Driver San Francisco which looks realistic enough) also makes a kind of farting sound (similar to motorcycle) when on low gears. Presumably this is some technical peculiarity absent in other cars. Why is that? 212.180.235.46 (talk) 09:58, 31 October 2017 (UTC)[reply]

Lamborghini makes a wide variety of cars, so I don't know which specific model you are referring to. There's a lot of general discussion here on the matter. --Jayron32 10:45, 31 October 2017 (UTC)[reply]

It's probably a V10 engine. The firing pulses from this are (unlike a V8) distributed more evenly so they don't have the typical V8 syncopated "burble". The silencing is also designed to cope with a loud engine at full rpm and so aren't as effective at silencing at low rpm, so there is more farting. Andy Dingley (talk) 10:55, 31 October 2017 (UTC)[reply]

See muffler. Also, we shouldn't make the assumption that all owners want their vehicles to be as quiet as possible. Making lots of noise may actually be the goal for some, and the manufacturers are aware of this. Of course, they can't make so much noise as to be illegal to operate. StuRat (talk) 15:02, 31 October 2017 (UTC)[reply]

I have asked this question at the Teahouse, so that it wouldn't have to be discussed at both articles' talk pages. In a nutshell, there are two articles that don't agree about what the Potrzebie date for October 27, 2007 is. -- 2601:381:101:142:51A:6609:6BDB:C397 (talk) 15:40, 31 October 2017 (UTC)[reply]

According to the Wikipedia article titled Potrzebie you can use Wolfram Alpha to perform the correct conversions for you. --Jayron32 15:43, 31 October 2017 (UTC)[reply]

By the way, I didn't realize I wasn't logged in until after moving the discussion here. Sorry about that. -- MrHumanPersonGuy (talk) 15:47, 31 October 2017 (UTC)[reply]

Went to the WA website, but I didn't see any way of converting the date to this format. Secondly, I can't find any real 'useful' entries on Google to assist in computing it manually. :/ - NsTaGaTr ^(Talk) 16:13, 31 October 2017 (UTC)[reply]

Saving you the trouble of digging out an old Mad from your attic, The "Coding Horror" blog post "The Enduring Art of Computer Programming", has the information you'd need to calculate it manually. Have fun, Knuth didn't design it to be easy. ApLundell (talk) 17:42, 31 October 2017 (UTC)[reply]

Using Excel, I've figured out the calculation for this (*and it's actually quite simple*). Every 100 days from October 1, 1952 is a "year", and every ten days is a "month". Both articles reference October 29, 2007, which should in fact be "Cal 7, 201 C.M." Today's date would then be "A 2, 237 C.M." :) - NsTaGaTr ^(Talk) 18:28, 31 October 2017 (UTC)[reply]

@MrHumanPersonGuy: Hope this helps :) - NsTaGaTr ^(Talk) 18:47, 31 October 2017 (UTC)[reply]

This sounds like Mad's equivalent of Notary Sojac. ←Baseball Bugs ^{What's up, Doc?} carrots→ 21:30, 31 October 2017 (UTC)[reply]

So the list of humorous units of measurement says that the Potrzebie date is based on the "clarke", which it defines as the "average earth rotation". Taken literally, that would not be a solar day (the usual 24-hour day), but the slightly shorter sidereal day. Is it clear which is meant? --Trovatore (talk) 21:46, 31 October 2017 (UTC)[reply]

In fact, it doesn't -- but since the "10 days per month, 10 months per year" conversion lined up perfectly with what they used as a reference point (*Oct 29, 2007*), I'm going with solar day instead of sidereal day. - NsTaGaTr ^(Talk) 22:02, 31 October 2017 (UTC)[reply]

I don't have access to Excel, but I do have access to Julian day#Variants. which tells me that today's JD is 2 458 058. We'll lop off the first two digits and call it 58,058. Yesterday's Potrzebie date was 2, so the last digit of the Potrzebie date can be found from the JD by adding or subtracting 5 and vice versa. JD for 1 October 2017 was 58,027 and for 1 October 2016 57,662. For 1 October 1952 it was 57 662 - (16 x 1461) = 34206. For 29 October 2017 it was 58,055 and for 29 October 2015 58055-365-366 = 57,324. For 29 October 2007 it was 57324 - (1461x2) = 54,392. Now there are (54392-34206) = 20,186 days elapsed, which is 201 years, 8 mingos and 6 clarkes. The date should therefore be clarke 7 of mingo Jug. of year 201, but that's not what is given above. Something is awry.

If we look at yesterday's date, the elapsed period is 58057-34206 = 23851, which converts into 238 years, 5 mingos and 1 clarke, giving a Potrzebie date of clarke 2 of mingo Hum. of year 239. Again, there is no correspondence with what is given above. I am totally bemused. 80.5.88.70 (talk) 11:38, 1 November 2017 (UTC)[reply]

I'm not sure why Julian date is being included in the calculations for a completely different system. If Oct 1, 1952 is day 1, then counting forward in time puts Oct 31, 2017 as day 23772. If 100 days comes out to a year, then it's been 237 years, with a remainder of 72 days. With 10 day months, that puts us into the 7th month, with another remainder of 2 days. Since there is the option for a month 0 (*in the tens place*), we need to add one to the month, resulting in 8, so the true date for yesterday would have been "A 2, 237 C.M." - NsTaGaTr ^(Talk) 13:48, 1 November 2017 (UTC)[reply]

Well, if 1 October 1952 is day 1, 1 October 2008 is day (1+(14x1461) = day 20455, 1 October 2007 is day 20089 and 29 October 2007 is day 20117. 29 October 2015 is 20117+(1461x2)=23039. 29 October 2017 is 23039+366+365=23770. My mistake was working out the JD of 1 October 1952 as 34206 instead of 34286 and 29 October 2007 as 54392 instead of 54402. Since the year is 201 rather than 202 the calculation obviously uses elapsed, rather than current years. For yesterday, if you're adding 237y 7m 2d, using elapsed years, counting the first month as 0 you should arrive at A 3 rather than A 2. 80.5.88.70 (talk) 21:18, 1 November 2017 (UTC)[reply]

As you were, it's 1 November which is A 3, so 31 October 2017 is A 2, as NsT said. However, his statement that

there is the option for a month 0 (*in the tens place*)

implies that the calendar uses elapsed months as well as elapsed years. No calendar uses elapsed months (the only thing that comes near it is baby ages) and elapsed years aren't used outside India. So, unless NsT knows something about this calendar that the rest of us don't I am inclined to think that he is wrong.

The only thing I know about Excel is that it thinks 1900 was a leap year. The day count for 31 October gone (23772) gives a date of clarke 2, mingo A, year 238 by my reckoning. NsT doesn't give the day count for 29 October 2007. I think it's 201117, and I think its Potrzebie date is clarke 7, mingo Cal., year 202. 80.5.88.70 (talk) 09:07, 3 November 2017 (UTC)[reply]

Excel was written to think 1900 is a leap year for compatibility with Lotus 1-2-3, which was under the same misapprehension. See Leap year bug. AndrewWTaylor (talk) 18:42, 3 November 2017 (UTC)[reply]

I can't seem to find anything in Wikipedia about the F29Y2K bug. As I understand it, some software was programmed to deal correctly with years divisible by 100 but not by 400 (which are not leap years), but was not correct for years divisible by 400 (which are leap years). I'm surprised we don't have anything, which makes me wonder if I was misinformed. But I thought this was an amazing fail — if it happened, it means the programmers had to actively add a "fix" that would not have any advantage until the year 2100, by which point the code would almost certainly no longer be in use, but that caused an error in the then-fast-approaching year 2000. Whereas if they had simply left it alone, everything would have been fine for more than a century. --Trovatore (talk) 18:53, 3 November 2017 (UTC)[reply]

It's mentioned in Year 2000 problem#Leap years (also in the lead) but not by the name you mentioned. It does links to a ref [1] which includes some examples although it's likely to be difficult to work out where they came from by now (if it was ever possible). It's suggested in both our article and the source that the reason may have been due to misunderstanding of leap year rules. Nil Einne (talk) 03:21, 4 November 2017 (UTC)[reply]

I don't follow the logic here. If the program was "not correct for years divisible by 400" I fail see how the fix "would not have any advantage until the year 2100" - 2000 is exactly divisible by 400 so it would have an advantage in 2000. One urban legend mentioned in the link (possibly by analogy with the fact that years which are exactly divisible by four are not normally leap years if they end in two zeros) is that years which are exactly divisible by 400 are not leap years if they end in three zeros. That's less accurate than what we have now, although there's an interesting proposal at User:Megalibrarygirl/Jacinto Quirarte (currently only viewable by administrators) which reduces the discrepancy to zero. 62.30.14.156 (talk) 15:03, 4 November 2017 (UTC)[reply]

Maybe I wasn't completely clear; in it's easier to restate than to re-read what I wrote and try to figure out whether I was clear.

• The original programs, before the fix, assumed that all years divisible by 4 are leap years. That was not correct, but the first time it would have mattered that it was not correct was 2100.
• The "fix" added the observation that years divisible by 100 are (usually) not leap years. That made the program correct for 2100, but incorrect for 2000, because they failed to include the exception-to-the-exception for years divisible by 400.

Hope it's clear now. --Trovatore (talk) 19:18, 4 November 2017 (UTC)[reply]

Yes the fact that programs which just assume all years divisible by 4 were leaps years are generally fine until they are likely to have been replaced, and this was only a problem for programs which tried to, but failed to properly account for the 100 year rules is mentioned in both our article and the ref. (Although I'm not sure how many of the cases were later fixes as opposed to being written like that in the first place. But you could still call this an attempted fix of one kind.) Nil Einne (talk) 08:08, 5 November 2017 (UTC)[reply]

The redlinked article above proposes to do away with all these fixes and replace them with a simple rule which will last forever. It's been discussed before Wikipedia:Reference desk/Archives/Science/2017 March 23#How accurate is this table? (March 17). 86.155.148.59 (talk) 19:37, 5 November 2017 (UTC)[reply]