Computing desk
< February 9	<< Jan \| February \| Mar >>	February 11 >

Welcome to the Wikipedia Computing Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

February 10

Statistically intractable?

Does something like a "statistically intractable" set of problems exist? That is, problems that can be solved in polynomial time with a non-statistically approach, but cannot be tackled with stats?--Terurme (talk) 12:36, 10 February 2017 (UTC)[reply]

That depends on who you talk to. In my opinion, you will find some on Project Euler. Many of the problems there are trivial to solve with a set of for loops, but extremely difficult to solve using statistics or, more importantly, calculus. 71.85.51.150 (talk) 13:00, 10 February 2017 (UTC)[reply]

I know Project Euler, and it is clear that many problems cannot be dealt with brute force, but they can be easily solved with the right algorithm. However, I didn't know that some cannot be solved with calculus at all. Do you know, by any chance, the number of the problems that fall into this category?

Has anyone an example of a problem that can't be dealt with calculus? --Terurme (talk) 13:28, 10 February 2017 (UTC)[reply]

This is a common example: You have a steel corrugated roof. From the side, it forms a perfect sine wave with a period of 'd' (from y=0 to y=0 is 'd', not from y=0, skip the next y=0, and then on to the next y=0. In other words, if you consider each hump of the sine wave to half a circle, the diameter is 'd'). You have a roof that is length 'l'. You have a massive flat sheet of steel. What length of steel is required such that when corrugated the steel will be exactly 'l' wide. This is *very* easy with a for loop. It is deceptively difficult with calculus.

Another one... You have a triangle defined by 3 (x,y) points. You have three circles defined by r1, r2, and r3. Can the three circles fit inside the triangle without overlapping. Circles and sides of the triangle can touch, but not overlap. Again, very easy to program. 71.85.51.150 (talk) 14:31, 10 February 2017 (UTC)[reply]

It is deceptively difficult with calculus. In your example, the period would actually be 2d if d were the diameter of a circle which followed the curve of each peak or valley, because it takes a peak and valley pair to make 1 period. But you said a period of d, so I'll go with that. So, assuming the semicircular shape you seemed to be assuming (and which is common for corrugated steel) for the sine wave produced in this way, the answer is actually fairly trivial to calculate using basic algebra: S = 1/2dΠ(l/d) where S is the length of steel required. That being said, sine waves don't generally form perfect semicircles out of their peaks and valleys so one is forced to do a lot more work to model a realistic sine wave. And yes, this is (to the best of my knowledge) very difficult to do in calculus, but very easy to do with a loop.

ᛗᛁᛟᛚᚾᛁᚱPants Tell me all about it.

15:07, 10 February 2017 (UTC)[reply]

The trick is getting that specific arc length on the last semicircle. If it comes out to a perfect multiple of d, not problem. If it is a fraction of d, you need to calculate the remaining arc length. That is why d is given as the diameter of the circle. You will use it to divide l to get the fraction left over and use it as diameter to calculate arc length. (And I did purposely make the problem much easier. In the original, the width of each semicircle was d and the height was h - so they weren't actually semicircles. They were ellipses. I found that complication to be silly.) 71.85.51.150 (talk) 15:20, 10 February 2017 (UTC)[reply]

The trick is getting that specific arc length on the last semicircle. If you're saying that l/d must be an integer, then yes; it's very tricky and requires calculus. If you're saying the last semicircle is going to have a diameter of l%d (note this is programming notation, % is the modulus operator), then it's a little tricky but not too bad (still better to do with a loop, but fairly easy with algebra).

And I did purposely make the problem much easier. In the original, the width of each semicircle was d and the height was h - so they weren't actually semicircles. They were ellipses. I found that complication to be silly. I figured it was on purpose, to make it easier to explain. It's pretty rare for the frequency and amplitude of a wave to be exactly the same. The answer I provided only works when that is the case, naturally. I think I could work out how to do it algebraically, but despite being a programmer, I've never been very good at math and I don't want to make my brain hurt. And I'd probably get it wrong, anyways (note how I had to correct my formula above from d/l to l/d).

ᛗᛁᛟᛚᚾᛁᚱPants Tell me all about it.

15:43, 10 February 2017 (UTC)[reply]

xxx.yyy is not responding. Recover webpage

I have Windows 10 and Microsoft Edge and recently went through a major software upgrade with a message saying to restart my computer and wait approximately 90 minutes for the updates to take effect. All I know about the change is that my slow Internet has gotten faster, though I made no changes that should have made that happen. I turn off the computer these days with the button that tells it to sleep, so it comes on in the middle of the night. Apparently an update of some kind has taken place.

Yesterday it seemed like every web site I went to gave me "yahoo.com is not responding" with a button to "recover webpage". If I clicked on the button I got a message with a green line moving back and forth saying Windows was searching for a solution. If they found one, they didn't tell me. Most of the time the problem cleared up quickly and I never had to click on the "recover webpage" button or anything else.

Most sites work--eventually. Some are faster than others. Some never give me the message. McClatchy Newspapers sites sometimes won't let me go to articles, but I found a workaround by right-clicking and copying the link and then pasting. With these sites, the page will disappear and then come back. tunegenie.com will update the playlist but I can't listen to the radio station; it never stops "buffering" and the blue circle just keeps on turning.

I do have some hints. I signed in to radioinsight.com a few minutes ago and got some javascript error. It didn't stop me from signing in but I was never able to "agree to terms", which was something I had to do today for some reason. I also could not ask a new question here by clicking on the ask a question button.— Vchimpanzee • talk • contributions • 18:15, 10 February 2017 (UTC)[reply]

That is a multi-part question. I cannot answer it all, but a possible reason that [your] slow Internet has gotten faster, though [you] made no changes is that the browser (Edge) was updated and in the process the browser cache was cleared; in which case, the connection has not really gotten faster, but the browser has. This definitely happened to me with a previous version of Firefox, although there is no real reason that cache-clearing improves your situation, even with a large cache size. Tigraan^{Click here to contact me} 19:01, 10 February 2017 (UTC)[reply]
this'll fix you right up. ;) $ᛗᛁᛟᛚᚾᛁᚱPants Tell me all about it.$ 19:25, 10 February 2017 (UTC)[reply]

I have said it many times but I am not changing browsers. Regardless of what you think, it's just too complicated.— Vchimpanzee • talk • contributions • 19:36, 10 February 2017 (UTC)[reply]

Well, you never said it where I've seen it. And "it's complicated" is just... well... wrong. You download, you double click, then you click okay a few times, then you're done. That's not complicated. Maybe you meant your reasons for not changing are complicated? Either way, if you want to troubleshoot the problem thoroughly, one of the earlier steps should be trying to access those pages with a different browser on the same computer.

ᛗᛁᛟᛚᚾᛁᚱPants Tell me all about it.

19:40, 10 February 2017 (UTC)[reply]

"it's complicated" is just... well... right. I have no idea what I'm getting into. At libraries, people are paid to deal with Firefox and its endless updates and endless problems. I'm here by myself and Edge has mostly been working fine. It was working fine earlier this week.— Vchimpanzee • talk • contributions • 19:43, 10 February 2017 (UTC)[reply]

An update: the ask a new question button works.— Vchimpanzee • talk • contributions • 19:45, 10 February 2017 (UTC)[reply]

Well, that's good.

Firefox and its endless updates and endless problems That's curious to me, because I've used it for over a decade, and never noticed it to be more problematic than other browsers. Indeed, I've noticed it to be significantly less problem prone than others. Also, all updates are automatic.If you're worried about breaking something during the install, don't. It literally takes just one button click to install it. Web browsers can be very complicated bits of software, but they're designed to be very modular and work across platforms. There are other options as well, such as Chrome, Safari or Dolphin.

ᛗᛁᛟᛚᚾᛁᚱPants Tell me all about it.

20:00, 10 February 2017 (UTC)[reply]

I know what I'm talking about. At libraries where I've used Firefox it is constantly causing problems. Partly because there are so many updates. You seem to have the idea that any update will be easy for me. In the first place, downloading updates has been a problem since I got this computer. So I gave up on that. When it happened without any problems, that was fine. In the second place, each update causes new problems. The library pays someone to fix them.— Vchimpanzee • talk • contributions • 20:26, 10 February 2017 (UTC)[reply]

Interesting problem... I would suspect you windows, your browser or your ISP. Because that's rather vague, and in an attempt to further pinpoint the problem, I would suggest to install another browser (firefox, chrome or whatever) and see what that does. You can always uninstall it later.

Another approach would be to take your computer to a friends home and try another ISP.

Also, you could try a friends computer at your internet connection.

Trying another OS on your PC would be nice too, but it's too much trouble.

Get the idea? Try to pinpoint the problem further down, and please report back here.

Well, perhaps another user recognizes the problem and has a faster approach. :) Jahoe (talk) 20:00, 10 February 2017 (UTC)[reply]

I'm not going to do any of those things because it's too much trouble.— Vchimpanzee • talk • contributions • 20:26, 10 February 2017 (UTC)[reply]

Not to be a dick, but if it's too much trouble to follow any troubleshooting steps, your only solution is to live with the problem or pay someone to fix it for you. Coming here was just a waste of time.

ᛗᛁᛟᛚᚾᛁᚱPants Tell me all about it.

20:29, 10 February 2017 (UTC)[reply]

I can't just take this computer to someone's house. This requires unhooking everything (and it was hard enough to hook it up) and carrying this heavy tower around. I don't know why everyone acts like it's easy to do the things they expect me to do.— Vchimpanzee • talk • contributions • 20:43, 10 February 2017 (UTC)[reply]

Well, moving a computer might be hard, but installing Firefox is trivial. I can't imagine what kind of problems your library is having with it. Perhaps they have some parental protection software installed that interacts badly with it, but I've used Firefox for many years and have never had a single issue with it. In any case, the troubleshooting you're being asked to do is just install Firefox and try it for a few minutes to see if your problems persist in another browser. You don't have to use it after that if you're still frightened of it. You can even uninstall it. But I will reiterate what's been said: if you refuse to do any troubleshooting because it's "too much trouble", there's not much more that anyone can do to help you. CodeTalker (talk) 20:58, 10 February 2017 (UTC)[reply]

If you can't access this website from your home computer, use a friend's. That's pretty much where you're at right now. There is no solution but paying someone to come fix it. I'm sorry, I really do want to help, but the conditions you're putting on that help make it impossible. I've done remote troubleshooting before, but to be perfectly honest, I would never let a random wikipedian connect remotely to my computer, and I wouldn't suggest you do so, either.

ᛗᛁᛟᛚᚾᛁᚱPants Tell me all about it.

21:07, 10 February 2017 (UTC)[reply]

Is there not some way when "Windows is searching for a solution" to tell me what it found? A few weeks ago I got a message about the update but don't know how to get to it now. It's not like it's in email.

There is a place to click to find out what updates I need. I clicked on that and now the updates are downloading. And I was just told to restart. I'll do it later, when I finish.— Vchimpanzee • talk • contributions • 21:27, 10 February 2017 (UTC)[reply]

If Windows had actually found a solution, it would have told you. Usually what's happening when you see the "Windows is searching for a solution message" is Windows Error Reporting is just uploading crash data to a Microsoft database for later analysis (or not). In my experience and from what I've read, it's possible but pretty rare for it to actually report a solution on the spot. CodeTalker (talk) 22:35, 10 February 2017 (UTC)[reply]

Is 10.0.14393 useful information?— Vchimpanzee • talk • contributions • 21:33, 10 February 2017 (UTC)[reply]

I have regular problems with incomplete web pages when my internet connection is slow. The symptoms are similar to those that you describe. If update 10.0.14393 is causing problems, then there is some (complicated) advice here, but I think it would be better to let updates finish. I hope that things sort themselves out on their own for you when the updates are complete. Dbfirs 21:35, 10 February 2017 (UTC)[reply]

Okay, thanks. I have a button to click on when I'm ready to restart. I'm going to the Help Desk and the Teahouse (to look for unanswered questions) and when I finish there, I'll click on that.— Vchimpanzee • talk • contributions • 21:43, 10 February 2017 (UTC)[reply]

And when I get back tomorrow, we will know if this worked.— Vchimpanzee • talk • contributions • 21:55, 10 February 2017 (UTC)[reply]

Sigh... Perhaps we should close this section here. If the OP doesn't want to go through any trouble this helpdesk can't do anything. I suggest the OP brings his computer anywhere and have it fixed for a fee. That's probably not even that expensive... Jahoe (talk) 23:55, 10 February 2017 (UTC)[reply]

If you look at the years of complaints from vchimpanzee, he doesn't accept any advice of any kind. He just complains to complain. The only solution is to treat him as a troll and completely ignore him. — Preceding unsigned comment added by 2600:1004:B154:F431:8DD3:F7E2:151D:5901 (talk) 12:19, 11 February 2017 (UTC)[reply]

I think that's a bit harsh, anon IPv6 editor. What the OP needs is a friend who lives nearby who can sort out his problems. Anyway, if I'm correct, they will eventually sort themselves out. Dbfirs 12:30, 11 February 2017 (UTC)[reply]

It's not that I don't accept advice. It's that the advice is too difficult for me to deal with. This morning after the restart, the problems have all gone away.— Vchimpanzee • talk • contributions • 16:50, 11 February 2017 (UTC)[reply]

Resolved

Thanks for reporting that back. We can all be happy now. :) Jahoe (talk) 17:05, 11 February 2017 (UTC)[reply]

Request from TheSandDoctor

Hi there, TheSandDoctor recently used a {{help me}} template to ask for help on the Hollywood Pinball talk page, but I told them it would probably be best to ask their question here, so I'm posting a little message here pointing towards TheSandDoctor's message. I'll post it below to so you can see it here:

I was unable to find any credits for producer, programmers, designers, artists, etc. for Hollywood Pinball. If anyone could help locate them, be my guest and feel free to add them. Thanks!
— — Preceding text originally posted on Talk:Hollywood Pinball (diff) by TheSandDoctor (talk⋅contribs) 23:50, 9 February 2017‎ (UTC)

If you can, please try and help them - if you can. Thanks! Seagull123 Φ 19:51, 10 February 2017 (UTC)[reply]