Wikipedia:Reference desk/Archives/Computing/2014 March 1

= March 1 =

Scrolling and highlighting no longer work
My computer, running (I think?) Windows 7, simultaneously changed two behaviors: What's going on? Michael Hardy (talk) 03:11, 1 March 2014 (UTC)
 * I can no longer move a scroll bar smoothly up or down; rather it makes large discrete jumps.
 * I can no longer highlight text in order to copy and paste.


 * I don't know, but it might be malware. Also, a reboot often clears out problems.  Bubba73 You talkin' to me? 03:34, 1 March 2014 (UTC)
 * Could be a problem with the mouse. Try plugging another one in, check connections.--Salix alba (talk): 16:20, 1 March 2014 (UTC)
 * This happened with one web site, but then it worked normally with others. Later it worked normally with that one again. Michael Hardy (talk) 05:11, 3 March 2014 (UTC)

People you may know
On what basis does facebook make suggestion and feature People you may know work.What is the algorithm and logic behind it.Can a user can analyse and infer some meaningful conclusion about who is visiting or intersted in him from this list. Even if no user data is provided like work education then also this feature works.As facebook does not divulge anything in this regard so i am making this query here.Also is it possible to use any indirect means or evidence to infer who has visited a users profile page by the user himself.117.194.246.24 (talk) 11:51, 1 March 2014 (UTC)
 * As far as I can tell, all it does is look at the "friends" of your "friends". The more friends you have in common with somebody, the more likely you are to know them. Looie496 (talk) 14:47, 1 March 2014 (UTC)


 * As far as I know, Facebook does not allow any user A to see what other users B may have viewed A's profile page. There are several scammy ads/sites that claim to do this, but I recall Facebook has basically come out and said they were scams and not to be trusted (I can't easily find this statement at the moment...). As for inferring their secret algorithms: the only way to do test this scientifically would be to create several accounts yourself, that are only linked to eachother. That way, you know the entire geometry of that network, and can make inferences based on each account's "you may know" results. Ideally, you'd create several such isolated networks of profiles, and use statistics to do hypothesis testing on specific features of the FB algorithm. However, FB would probably frown on that approach, and may consider it a violation of their terms of service. SemanticMantis (talk) 19:01, 2 March 2014 (UTC)


 * Every time it has made a suggestion to me, it listed the number of friends we have in common. I see no reason to believe the algorithm goes any farther than that. Looie496 (talk) 16:41, 3 March 2014 (UTC)
 * User:Looie496, I agree that number of friends in common must be a big part, and it does list that number clearly. But here's my reasoning that leads me to believe that the algorithm is not just a simple list ranked by number of friends in common: person X shows up in my "you may know" list only sometimes. Other times, X may show up first on the list, other times 4th place on the list, even though the number of friends in common hasn't changed. The fact that the order and composition of the list changes suggests to me that there's something more going on in the algorithm (again, with no changes in number of friends in common). For example, suppose X has three friends in common with me (A,B,C), and Y has three friends in common with me (D,E,F). I suspect that X will be promoted when I have posted on A's wall recently, while Y will be promoted if I have recently interacted repeatedly with D and E. Stu's experience below also indicates that friends in common is not the only criteria. Of course, I can't know for sure, that's why I suggested an experimental approach. SemanticMantis (talk) 18:02, 4 March 2014 (UTC)


 * I don't use Facebook, I just signed up once for some promotion, and never "friended" anyone or signed back in. However, it still suggests several people to me each day by email who "I may know".  I don't know any of them.  However, I live in Detroit, which is mostly black, and it keeps suggesting names like Shaniqua, which sure sound black to me, so I have to think it's going off geography alone.  Perhaps it uses this as a backup method for those who haven't friended anyone. StuRat (talk) 16:56, 3 March 2014 (UTC)

What's wrong with this PHP snippet?
I'm trying to teach myself PHP etc. from the book PHP, MySQL, JavaScript, & CSS from O'Reilly. I got through setting up a server on my MacBook and tried out the first little bit of PHP that they use as an example:

It's in an HTML test page along with a simple JavaScript test which works fine. This line shows up as "Hello World! Today is . How are you?" So, where did the date go? I typed it in exactly as the book has it. Thanks for any help, Dismas |(talk) 22:00, 1 March 2014 (UTC)


 * I've figured out half of my problem. I hadn't changed the file name from .html to .php.  But now it just says ..."Today is 1."...  What's the fix for that?  Dismas |(talk) 22:07, 1 March 2014 (UTC)


 * Got the rest of it. The font that the book is written in made me think the lowercase letter "l" was a numeral "1".  Got it now.  Dismas |(talk) 22:12, 1 March 2014 (UTC)

Dismal determined that nothing wrong with PHP; Problem was between his monitor and chair. 203.77.233.158 (talk) 13:35, 2 March 2014 (UTC)
 * Thank you for your snarky comment and misspelling my name. Don't quit your day job. Dismas |(talk) 14:01, 2 March 2014 (UTC)


 * Some have argued, though, that there is everything wrong with PHP . → Σ σ  ς . (Sigma) 08:07, 3 March 2014 (UTC)