Wikipedia:Reference desk/Archives/Miscellaneous/2014 June 16

= June 16 =

Facebook friend suggestions
Hi, I know this is about Facebook, so it could be in the Computing helpdesk, but this isn't so computing related. It's half computing, half sociology... I think.

How does Facebook suggest friends? I showed my mum how to use the site a couple of years ago, and he was amazed how it could suggest people she hadn't seen or spoken with in ages. At the time, I said it was the genius of the Facebook algorithm, and Facebook worked it out from her current friends and their friends.

However, I have just made a new account (I'm sick of the old one), and without having added any friends, or searched for anyone, or anything like that, Facebook is suggesting me friends from my primary school, and secondary school. Some of these names I haven't seen or heard since those days, so how has it figured out I once went to school with them? It can't just be from me using the same name, because it hasn't suggested the good friends that I have now and contact on Facebook every day with my previous account. It can't be my IP address; my IP address is from my university house, which is half a country away from my primary and secondary school. So how does it know!? It's majorly creepy!

81.110.73.68 (talk) 04:18, 16 June 2014 (UTC)
 * Facebook can use any information you give; the schools and collages/universities you attended (together with your age), where you worked, and what clubs you belong to, and friends-of-friends. If you allow it, FB will read your email provider's address book, and suggest everyone on it. If someone else has allowed FB to read their contact book, and you're in it, you might get friends suggestions for them. CS Miller (talk) 08:02, 16 June 2014 (UTC)
 * The experience of 81.110.73.68 is very similar to my own. I had done very little - enter name and current home town, but had not linked any schools, colleges etc. I definitely didn't release my email contacts. I found three people that I was in regular contact with, and sent them friend requests. Facebook then came back with a long list of friend suggestions, and 15 or so of the first 20 names on the list were all people that I was at primary school with - before 1976. That school is about 30 miles from where I live now, and was not attended by those three people to whom I had sent friend reqs. -- Red rose64 (talk) 16:19, 16 June 2014 (UTC)
 * The thing is that when you disconnected from FB, killed your account and started a new one - you expected FB to not be able to figure out who you are...and that's just not true.
 * They know what IP address you used to work from - and they new which IP address this seemingly new user uses...now they may kinda suspect that this new user is the same as that old one that was recently cancelled. Even if you disconnected your web interface and had your ISP provide a new IP address, they can still do a reasonable job of "geolocation" to figure out which city you're in.
 * If you didn't delete all of the Facebook-created cookies (and all of your flash cookies and all of the "Local Storage Objects" in your browser then they immediately know who you are - irrespective of your login account.
 * Even if you very carefully cleaned up every scrap of those things, they can compare a bunch of things that the browser tells them about your computer - which browser revision, which operating system revision, perhaps the capabilities (and therefore the type) of your graphics card, how fast your CPU is.
 * Using JavaScript, it can figure out a lot of the options you have set - what fonts you have installed - what screen resolution you have.
 * It can tell which website you were using when you decided to head over to Facebook (the "referrer").
 * It can store your name in the RGB values of an image file. Your browser will cache it and not bother to reload it later if you go back to the same site.  JavaScript can look at the pixels of the image and figure out what name you used last time you logged in to that page.   Stopping it from doing that requires that you turn off web caching altogether...that would slow your browsing experience down to a snails' pace!
 * There are a huge number of tiny pieces of information that are leaked about you through your computer.  Obviously, putting together all of these circumstantial things *might* be wrong - but there a re a lot of them.  If it can ask 100 questions and get answers to just 50 of them - then it can probably figure out exactly who you are - even if the other 50 don't match from the last time you visited FB.
 * They have nothing to lose here. If they guessed wrong, then all that would happen is that it might make some bad suggestions.   But if it gets it right, then it seems pretty magical.   Every piece of information you give - even in the tiniest of ways - is a clue.  If you re-friend just one of the people on the suggestions list - then it's pretty much certain to have guessed right - and it knows that.   It can produce a confidence figure that climbs rapidly to 100%.
 * You might ask why they go to all this trouble - surely just offering you a better suggested friend list isn't worth all of that trouble? The real reason is that they are selling all of this information to their customers so they can target you with adverts that you're more likely to respond to or figure out what percentage of 22 year old male hispanics watched such-and-such TV show and also like to eat at Joes' Diner.   If you were able to completely flush your old identity - then all the information they have becomes worthless - but if they can preserve it, even though you switched accounts, then they've protected their investment in giving you access to this software, disk space and server time.   So you can expect them to work hard to keep on knowing who you are...and it's increasingly difficult to hide.
 * Scarily, this isn't hard to do. These are called (variously) server-side cookies, forever-cookies and so forth.  This site http://samy.pl/evercookie/ offers a JavaScript module - and I'm sure that commercial software is much, much better at it.
 * You might think that your setup is pretty standard - so there aren't many things to track you by....but try https://panopticlick.eff.org - it has a database of some 4 million people and their computers.  The question it asks is how unique you are amongst those 4 million users - how many of them could be confused with you?   The answer in my case is none.  My computer is sufficiently unique that by testing the handful of things it tests, they can figure out who I am precisely within the 4 million people they have on file.  Probably, my computer, the settings on it and the other stuff I inadvertently leak is enough to tell exactly who I am, and probably even track me across the four different computers, one phone and two tablets that I use regularly.
 * Scary?....yes, definitely.  Avoidable?...not really, at least not if you need any kind of convenience at all in your browsing experience.
 * SteveBaker (talk) 18:37, 16 June 2014 (UTC)
 * I wasn't creating a second Facebook account. It was my very first: I cannot have had an "old identity". Also, my IP address is dynamic, and changes two or three times a week - but it always geolocates to 51.5°N, -0.13°W (that's the Methodist Central Hall Westminster for some reason), which is 79.13 km from where I really live. My old primary school is 98.33 km from that point, and 42.93 km from where I now live. Since getting internet-connected at home (some years earlier), I had never looked up that school, nor searched for anybody connected with it, so there cannot have been anything in my browsing habits to suggest that I was in any way remotely interested in getting in touch with people from 38 years ago. I may have had some Facebook cookies: but they will have related to friends and colleagues who I met several years after leaving school, and who now live in various odd corners of Britain. -- Red rose64 (talk) 22:04, 16 June 2014 (UTC)y
 * Well, if Facebook knows that you know those other people, perhaps it can deduce these old school buddies. The fact is that it did do it - and it must have gotten the information from something you've done in interacting with them.  Companies like this are very secretive about exactly how they do it...but it'll be something along those lines.  Something you've said or done in the past made that connection - probably something you weren't even aware was being logged.  If Facebook knows your real name, then it's possible that one of those people on the suggested-friends list talked about you or something.  Without knowing every interaction you ever had with those people and with Facebook, it's impossible for us to tell you exactly how they did that...but it wasn't telepathy - somehow, you told them enough to make the connection. SteveBaker (talk) 23:18, 16 June 2014 (UTC)
 * That's exactly what the deep-cover Chief Telepathy Officer of Facebook would say. --Trovatore (talk) 23:44, 16 June 2014 (UTC)

Thanks to everyone that responded, but SteveBaker you are the best!! Thanks! I haven't told Facebook on my mew account anything but my name and D.O.B. and so from there it must have connected me. I haven't yet deleted the old account, but I didn't bother to delete cookies, LSOs, or anything like that, so I guess all that sort of stuff must have helped them too. I guessed that they must realise that many people log into Facebook from the same computer as others, so that information wouldn't matter. I don't really mind them doing it though. Frankly, it makes it a little easier to add my friends, if they get it correct. But it's just really weird that just as I made my new account, it suggested people, and the people it suggested aren't people that I have ever been connected with on my old account, and I've never searched for some of them before (as they went to my primary school), but just from the name, D.O.B., and this background data they know that I once interacted with these people. It would make so much more sense if they suggested people that I am currently connected with, and interact with. Although, my current friend list is hidden from the public, so if they did that, then starting an account under someone else's name would be an easy way to get who they're closest friends with.

Thanks everyone! 81.110.73.68 (talk) 03:25, 17 June 2014 (UTC)

I just figured it out. Thanks CS Miller!! When I set up the account, Instead of making a new fake e-mail address for it like I normally would, I was lazy and used an old e-mail address. And these people might still have me in their address book, and they have let Facebook access their address book, so now Facebook suggests them to me. It makes so much sense now! 81.110.73.68 (talk) 03:29, 17 June 2014 (UTC)

Playboy and it's use of maximum age limits past to present?
Apparently, Playboy doesn't really have a maximum age limit on women that it considers attractive. However, I know that Playboy has used a maximum age limit on two occasions when it has searched for attractive women. When the first Playboy Club opened in Chicago in 1960, it ran this newspaper advertisement:

GREAT OPPORTUNITY FOR THE 30 MOST BEAUTIFUL GIRLS IN CHICAGOLAND

''Playboy is opening a new key club … catering to Chicago’s most prominent executives and sportsmen. To serve our exclusive clientele and decorate the club, we are looking for thirty single girls between 18 and 23. Experience is not necessary. Just be beautiful, charming and refined.''

Secondly, Playboy posted this notice for a casting call in 2012:

PLAYBOY CASTING CALL*~ Casting models for upcoming Playboy shoots. Attractive females, 5’0”-5’11" natural or enhanced, minimal tattoos, in shape, model types &girl next door. 18-28 yrs, all ethnicities. Model Rates: $500.00/day for Playboy.com shoots. Playboy conceptual/theme shoots $1,000.00/day $25,000 for Playmate & Opportunity to appear on “Playmates” Models selected will do a test shoot for Playboy Magazine & and will also get to be featured on Playboy TV’s series “Playmates”. If chosen to become Playmate, model will be awarded all Playmate rates. We will have Playboy shoots in New York for local models and will also fly models to Los Angeles for photo shoots. Playboy photo shoots are nude and very tasteful. To see photos go to: REDACTED or REDACTED We are holding a New York casting call on October 15,16 & 17. We also have weekly castings at our Playboy studios in LA. Submit photos, name, age, contact info & City you live in to: REDACTED Include the day you’re available for the casting. Job type: Paid Casting Location: New York Contact email: REDACTED

So I'm wondering, has Playboy used a maximum age limit on occasions other than the two I mentioned above? Hanna Nymous (talk) 10:07, 16 June 2014 (UTC)


 * Rebecca Ramos pre-dates the second posting you mention by about a decade. She was 35 when she was a Playmate.  And just two years before, Jaime Faith Edmondson was 31.  Dismas |(talk) 10:18, 16 June 2014 (UTC)
 * I'm sure Playboy makes some exceptions. After all, they wouldn't be able to use an ageless yet 34-year-old yet 22-year-old, 8'6", jaundiced women as their cover model otherwise, would they? — Crisco 1492 (talk) 10:27, 16 June 2014 (UTC)

Birthdate
I've been working on cleaning up and verifying the article on American artist Molly Crabapple. However, I'm currently stuck on verifying her birthdate, allegedly September 13, 1983. I believe that there are reliable sources that confirm her birth year, but I haven't found any that verify her birth month and day. I think I must be looking in the wrong place because I haven't had this much trouble before. Any ideas? She's a New York-based artist who also used to model. Viriditas (talk) 10:20, 16 June 2014 (UTC)
 * Some people are quite protective of this information. Bear in mind that you should only use dates of birth in Wikipedia that have been widely published by reliable sources. See WP:DOB.--Shantavira|feed me 12:54, 16 June 2014 (UTC)
 * Her website has an email address; you could contact her and ask.--Dreamahighway (talk) 19:22, 16 June 2014 (UTC)
 * No. Articles are based on published sources. We do not use personal communications as sources. AndyTheGrump (talk) 21:34, 16 June 2014 (UTC)


 * You could try it (being sure to let her know what you intend to do with it) and see what answer you get; and then qualify it by saying that "according to the subject" her birthdate is such-and-such. However, if she wanted her birth date and/or year widely know, she would have released that info already. ←Baseball Bugs What's up, Doc? carrots→ 00:04, 17 June 2014 (UTC)


 * If I found a birthdate "according to the subject" without a reference in a Wikipedia article, I would remove it as unacceptable. --ColinFine (talk) 08:18, 17 June 2014 (UTC)
 * This has her turning 28 in early September of 2011. Bear in mind that someone saying they're a certain age is not necessarily reliable. As an extreme example, Jack Benny died at the age of "39", despite appearing to be in his 80s. ←Baseball Bugs What's up, Doc? carrots→ 00:29, 17 June 2014 (UTC)
 * After further research, it appears that she has spoken about her birthday being on September 13, 1983, on multiple occasions on her Facebook page. Additionally, this birthdate appears on at least one professional site she uses to arrange lecture tours.  I think that's a good indication that the information is valid. Thanks. Viriditas (talk) 04:25, 17 June 2014 (UTC)
 * The Facebook mentions are irrelevant. The appearance in the professional site is probably acceptable, unless somebody were to challenge it. --ColinFine (talk) 08:18, 17 June 2014 (UTC)
 * I'm pretty sure they are relevant per WP:SELFPUB. Viriditas (talk) 01:00, 18 June 2014 (UTC)


 * The point of BLP rules is to prevent harm to the subject. If the subject publicly states they were born on such-and-such date, there's no harm from us citing that statement. Every allegedly "reliable published source" will likely have gotten that info directly from the subject anyway (especially if the subject was born after the 1940 census). There's no reason to play these kind of games here. ←Baseball Bugs What's up, Doc? carrots→ 02:46, 18 June 2014 (UTC)

Typing pool photo?
Half way down this page is an odd-looking photo that according to the image properties is of a typing pool in 1970. What are the women doing? What are all those screens and gadgets in front of them, and how do they know what to type on the teenyweeny keyboard in front of them? --Dweller (talk) 10:52, 16 June 2014 (UTC)
 * Producing punch cards for computers, maybe? The machines look prettly similar to the one in this picture. Deor (talk) 11:05, 16 June 2014 (UTC)
 * Yes, they are entering data on Hollerith punch cards.  → Michael J Ⓣ Ⓒ Ⓜ 12:12, 16 June 2014 (UTC)


 * Yep - I agree, those are punched card machines. I used one back in 1976 when I was learning computer programming.  The keyboards don't need all of the arrow keys or most of the punctuation symbols, shift/alt/control keys and so forth because punched cards were mostly uppercase only and with a very limited set of other symbols.  The three sections you can see at the back are probably the blank card hopper, the punching station in the center, and the finished card hopper on the other side.  You had the ability to duplicate sections of one card onto the next - so it keeps one card on the output side to copy from.  Each card has room for 80 characters in one line of text.  In addition to punching the holes, it also typed the letters along the top of the card so you could see what you'd just typed.   There were also features for automatically numbering the cards and employing tab-stops and such - and you could set up a "program card" on a drum inside the machine that let you do various other automatic functions.  I'm not sure about the screen above the keyboard.   I've never seen that.  It suggests that these are more recent machines than I ever used...probably at the very end of the era of punched cards. SteveBaker (talk) 17:23, 16 June 2014 (UTC)
 * One additional detail: in the link Deor posted, the program drum that Steve mentioned is visible through the little window at top center. They happen to have used a pink card for the program card.  To install it, you opened the window, lifted the drum off its mount, opened the clip holding the old one to the drum, bent the new card smoothly around the drum, and refastened the clip to hold it. --69.158.92.137 (talk) 04:29, 17 June 2014 (UTC)
 * I don't think those are screens above the keyboards. (Note that they are not apparent at some of the stations—such as the two behind the first two in the rightmost row.) I think they're punch cards or some other type of cards from which the typists are copying. Deor (talk) 17:37, 16 June 2014 (UTC)
 * Probably paper forms rather than some type of cards. --69.158.92.137 (talk) 04:35, 17 June 2014 (UTC)
 * Yes, when I started work in an insurance office in the 1970s, data was entered into a distant computer by us writing the figures onto a paper form with a ballpoint pen, a carbonless copy was sent off to a "computer bureau" (collected by van once a week and thence by train from London to Leeds where typist ladies were paid less), and we received a huge printout a week later showing what mistakes had been made - usually the typists had misread our writing. A typing pool was rather different; if you wanted a letter to be typed, you dictated it onto a Dictaphone (tape recorder) or scribbled it on a piece of paper, and somebody in a room full of typists turned it into the finished product. Our firm was too small to have a typing pool, unless you were a senior manager with your own secretary, you just had to talk very nicely to the receptionist. Lots of business letters ended up being hand written unless they were REALLY important. Alansplodge (talk) 21:45, 16 June 2014 (UTC)


 * Definitely keypunch machines, as SteveBaker and others have described. Several illustrations in Keypunch are similar. ←Baseball Bugs What's up, Doc? carrots→ 00:00, 17 June 2014 (UTC)
 * If you Google Image "ibm punch card machine", you will find many, many examples. This one is pretty good, as it labels the various parts. What looked like a "screen" in the OP's picture is actually just a card or some sort of hard-copy they would read from. Note that the OP's picture shows the keypunchers in pairs. I expect one is doing the initial keying and the other is verifying. It's been about 35 years since my company last used punch cards, but as I recall, the same machine could be used for both. That's where the function control switches come in. ←Baseball Bugs What's up, Doc? carrots→ 08:31, 17 June 2014 (UTC)
 * I think that they are the IBM 026 Keypunch and IBM 056 Verifier (introduced 1949, not superseded until 1964 and the IBM System 360 which needed more sophisticated punch cards). They would be paired up in use. -- Red rose64 (talk) 09:32, 17 June 2014 (UTC)
 * Thanks all. How do they know what they're supposed to be typing? They don't seem to have any headphones. Paper notes? --Dweller (talk) 13:08, 17 June 2014 (UTC)
 * Dictaphones would use headphones perhaps. Keypunchers were typically entering data, which would be pretty tedious to read out loud. ←Baseball Bugs What's up, Doc? carrots→ 14:39, 17 June 2014 (UTC)
 * Typically, there would be a sheet of paper (see those under the left hands of these women, you need to zoom right in; or the one at the centre of this photo) divided into a number of rows, each of 80 columns. The keypunchers will produce one card per row. -- Red rose64 (talk) 15:36, 17 June 2014 (UTC)