Computing desk
< April 19	<< Mar \| April \| May >>	April 21 >

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.

April 20[edit]

do people get their emails[edit]

I sent an email to someone at their work address, having no other contact details for the person, and they haven't replied. Do people at companies get their email, if sent from an address that is not listed in their address book, or is it always fairly sure they have received it? If they always get it (ie. with only the usual spam filtering), how do they deal with the spam that must come from having an email address listed publicly? Thanks in advance, It's been emotional (talk) 03:24, 20 April 2011 (UTC)[reply]

We can't possibly know. Maybe they got it. Maybe their company spam filter ate it. There's no generalized rule. Different companies use different spam filters. Different individual have different patterns of e-mail use. Wait a week, send the e-mail again with a "Hi, I'm not sure this got through..." and see if that helps. --Mr.98 (talk) 04:03, 20 April 2011 (UTC)[reply]

It would be pretty unusual for a company to dis-allow mail from unknown senders. (It would make it difficult to conduct business.)

It wouldn't surprise me if some company, somewhere has tried it, in the name of security.

It's very likely that the mail will be monitored by someone and an unusual email address might get their attention. Depending on how oppressive the workplace is about those sorts of things, the person you emailed might not want to even acknowledge the email until they could use a personal account, for fear of being accused of misusing company time and resources. APL (talk) 19:58, 20 April 2011 (UTC)[reply]

There's also nothing in email that guarantees delivery in any specific amount of time, it could take a week, it may have not even reached their servers at all. If you want to really talk to someone, email is not the way. ¦ Reisio (talk) 09:13, 20 April 2011 (UTC)[reply]

An email delivery that took a week would not break any rules. It would still be standards-compliant. ... But I think it's safe to say that if it took more than a couple of hours, most people would consider that a problem. APL (talk) 19:58, 20 April 2011 (UTC)[reply]

One of the best comments I've heard about email is that its reliability and security is worse than what you would expect with a postcard sent from another continent while on vacation. HiLo48 (talk) 23:57, 20 April 2011 (UTC)[reply]

That's probably true from a strictly autistic point of view, but a whole person knows that postcards are guaranteed to be handled by at least one person - your mail carrier - whereas most e-mails are never read by any human but the recipient. I'd rather e-mail a hundred friends my bank information than open a paypal account; and that doesn't mean that I would rather send a hundred friends post-cards with the same information from another continent while on vacation. Like it or not, e-mail is more like having a phone conversation over a wireless telephone (i.e. landline but with a wireless headset) that anyone can listen in on from across the street. Would you ever dictate your credit card number over such a phone? Obviously yes. 188.157.230.192 (talk) 10:13, 22 April 2011 (UTC)[reply]

"a whole person knows that postcards are guaranteed to be handled by at least one person"

And most people do not realize that email is entirely insecure and utterly without any guarantee of reliability, so as a way to put that information into terms people who have mailed postcards will understand, it's pretty good. ¦ Reisio (talk) 10:21, 23 April 2011 (UTC)[reply]

also the mail carrier does so under calm and inconspicuous conditions, with good reason to handle the postcard, and plausible deniability should they fail to deliver it to you (human error, e.g. misread name, whatever) along with many easy chances to misappropriate it afterward. In fact, you could even make a thief out of an honest carrier by presenting a real "crime of opportunity", if your postcard happened to include a very large credit card authorization, some multimillion dollar yacht or whatever. There is no real difference between sending an e-mail that probably a human is not going to read with your credit card information and a $30 subscription authorization, and the same e-mail for a yacht and the addition of a bunch of zeros to the sum; unlike a postcard, where this is a very real difference! In short: the analogy is not really applicable for a variety of reasons, and does not serve a useful purpose. 188.157.230.192 (talk) 10:20, 22 April 2011 (UTC)[reply]

Firefox 4 Navigation Bar[edit]

So I recently got Firefox 4, and it's all good, except for one thing. When you type something without a .com or .co.uk or whatever into the bar it used to try to find you that website, and if it failed would search it. So typing 'google' would get me google and typing 'BBC' would get me the news. But now it just does what the others do, searches for it in Google. So typing Google doesn't get me Google it gets me the Google of Google, and so on (I should have used a better example). Is there any way to change this back to what it used to do, since this is really bugging me now? Prokhorovka (talk) 13:37, 20 April 2011 (UTC)[reply]

Read this official guide from Mozilla to configure your Location Bar. This feature (treating the location-bar as a "smart toolbar" instead of a strict URL parser) is a very dangerous security vulnerability, because it is incredibly prone to phishing attacks; for this reason, Mozilla and other major web browser vendors are trying to "un-train" users who have grown dependent on it. Read more about Mozilla Phishing Protection design if you are interested in this topic. Nimur (talk) 18:07, 20 April 2011 (UTC)[reply]

The keyword.URL value you'd want would be something like this: http://www.google.com/search?ie=UTF-8&oe=UTF-8&btnI=&q= ¦ Reisio (talk) 22:12, 21 April 2011 (UTC)[reply]

Cheers, that's exactly what I was looking for. So it's dangerous to have the auto-complete, but not the auto-search you say? Good to know. Prokhorovka (talk) 18:14, 20 April 2011 (UTC)[reply]

Well, if you understand how to read a URL, there's very little that a phisher can do to "fool" you; so it doesn't matter what the browser delivers to you (you will be able to determine whether the delivered content is malicious or trustworthy). The problem is, most internet users have no idea how to read a URL, and tend to be very trusting of "whatever the browser returns," which could be a Google "I'm Feeling Lucky" search result, or could be anything else. Safe browsing will always require a little human-side intervention. The "danger" is breaking the one-to-one mapping between what you typed, and what you got; the "smart URL bar" does this for convenience. Nimur (talk) 18:17, 20 April 2011 (UTC)[reply]

As to how to teaching someone how to read a URL, some Carnegie Mellon researchers wrote a game that teaches you to discriminate phishing links from genuine. Their (regrettably rather minimal) demo is here now. -- Finlay McWalter ☻ Talk 18:27, 20 April 2011 (UTC)[reply]

That looks a little dangerous itself. Yes, a funny URI can tip you off, but a perfectly ordinary looking domain can have malicious intent, too, and what makes it phishing is when they go after your personal data, not just visiting the site. ¦ Reisio (talk) 22:12, 21 April 2011 (UTC)[reply]

I should be OK so long as I always type it in myself if I'm logging into my banking, look for https if possible and generally read properly any sensitive sites addresses right? Prokhorovka (talk) 18:31, 20 April 2011 (UTC)[reply]

Sure. I mean, phishing isn't going to break your computer - it just lulls you into a false sense of security. If you are careful about which sites you use when you enter your financial and personal details, you will generally be okay. Nimur (talk) 18:34, 20 April 2011 (UTC)[reply]

You should consider using keyword bookmarks instead. Bookmark every page that you want to access by keyword, and in the "keyword" field of the bookmark properties, enter something like "bbc" or "google", or even "b" or "g". Then typing the keyword into the address bar will immediately take you to the bookmarked page. Keywords can also take arguments. For example, if you bookmark https://encrypted.google.com/search?q=%s and give it the keyword "g", then typing "g keyword" in the URL bar will immediately take you to the search results for that keyword. You can attach a keyword to http://www.bbc.co.uk/search/news/?q=%s to search for BBC articles, https://encrypted.google.com/search?q=site:en.wikipedia.org+%s to search Wikipedia using Google, etc. -- BenRG (talk) 20:48, 20 April 2011 (UTC)[reply]

Then you're trusting your bookmarks database instead of your eyes. I'll take my eyes. ¦ Reisio (talk) 22:12, 21 April 2011 (UTC)[reply]

Huh? If you don't trust the keywords you created (why?), you can still look at the URL, which replaces the keyword in the address bar as soon as you press Enter. -- BenRG (talk) 22:26, 21 April 2011 (UTC)[reply]

Because anyone who has access to your computer for just a few minutes can change a bookmark keyword. Yes, you can still look at the address bar, if you haven't developed a habit of relying on your bookmarks being invulnerable. ¦ Reisio (talk) 22:50, 21 April 2011 (UTC)[reply]

But why would anyone do that? Do you really believe that phishers wander around modifying bookmarks on unattended computers? They don't, because the cost and risk would far outweigh the expected reward, but even if for some unclear reason they decided to do so to your computer, why wouldn't they just install a keylogger instead? -- BenRG (talk) 09:02, 23 April 2011 (UTC)[reply]

To exploit you. Phishers exploit people who aren't paying attention, people who click random links and assume the page that opens is what it appears to be, which is what most people do when they click (or otherwise open) a bookmark. Mmmm, and why not just rob a bank to get rich? I think you're underestimating the variety of exploitation people embrace. :p ¦ Reisio (talk) 10:25, 23 April 2011 (UTC)[reply]

Again: phishers do not go around modifying the bookmarks on people's computers. That threat exists only in your head. -- BenRG (talk) 04:34, 24 April 2011 (UTC)[reply]

All threats only exist in someone's head before they are actualized.* ¦ Reisio (talk) 15:20, 24 April 2011 (UTC)[reply]

Blender 3D Rendering Problems[edit]

I am able to render a movie in whatever format I would like to in Blender 3D 2.56, but only the "AVI RAW" format works in Windows Media Player. By that I mean the other formats do play, but the screen is black like if it weren´t a movie I tried to play? 83.183.172.203 (talk) 17:02, 20 April 2011 (UTC)[reply]

It sounds like you have a problem with Windows Media Player and installed codecs on your system that is unrelated to Blender. You may find the tool ffdshow helpful; it installs ffmpeg on Windows as a DirectShow filter, (and usually solves most common codec problems). Instructions and a tutorial are available here: FFDShow on VideoHelp.com. Nimur (talk) 17:36, 20 April 2011 (UTC)[reply]

Simpler and more useful to you in the long run to switch to VLC media player for all your media playback needs. ¦ Reisio (talk) 22:44, 21 April 2011 (UTC)[reply]

The mysteries of .NET appdomains[edit]

Today at work, I had the following problem:

The application I was working on creates several appdomains running concurrently. I was developing a piece of code that is used by each appdomain, but is the same regardless of which appdomain it's used by. It has an initialisation method that only has to be run once in the entire application, so I tried to make initialisation checking code to prevent it from being run multiple times. This turned out to be more difficult than I thought.

I first tried a simple class-level static flag and a monitor statement to check and set that flag. That failed miserably, because it appears each appdomain has its own private copy of the flag, which makes it useless. Then I discovered we already had a class that was intended to be shared across several appdomains, which inherited from MarshalByRefObject. I added the following code to it:

private Dictionary<string, object> attributes = new Dictionary<string, object>();
public Dictionary<string, object> Attributes { get { return attributes; } }

However, when I then tried to check and set flags via the Attributes property, this didn't work any better. I changed the implementation to:

private Dictionary<string, object> attributes = new Dictionary<string, object>();
public void SetAttribute(string name, object value) { attributes.Add(name, value); }
public object GetAttribute(name) { return attributes.ContainsKey(name) ? attributes[name] : null; }

And then it worked. But why did the first version not work when the second one did? JIP | Talk 19:07, 20 April 2011 (UTC)[reply]

I have no experience with this, but I imagine the dictionary is returned by value to the caller (since it doesn't inherit from MarshalByRefObject) and any subsequent changes by the caller affect only the copy. In the second case the dictionary stays on the callee's side and name, value and the return of GetAttribute are the only things that go over the wire (and they will be silently copied unless they inherit from MarshalByRefObject). -- BenRG (talk) 20:30, 20 April 2011 (UTC)[reply]

I showed this code, which I had written, to my boss:

SharedObject o = GetSharedObject();
lock (o) {
  if (o.GetAttribute("flag") != null) {
    return;
  }
  o.SetAttribute("flag", true);
}

and he said the code is not thread-safe. Why not? Is it because the statement lock (o) is in fact locking the appdomain's own reference to the object, not the object itself? Would it be better if I wrote it like this?

/* inside the shared object */
public bool CanEnter() {
  lock (this) {
    if (GetAttribute("foo") != null) {
      return false;
    }
    SetAttribute("foo", true);
    return true;
  }
}
/* in the calling method */
SharedObject o = GetSharedObject();
if (!o.CanEnter()) {
  return;
}

JIP | Talk 18:11, 21 April 2011 (UTC)[reply]

49 7/8' cable[edit]

I was doing a search for Ethernet cable on Best Buy's web site when I ran across this. It's 49-7/8' long. Why such an odd measurement? It's not even in meters either. Dismas|^(talk) 23:08, 20 April 2011 (UTC)[reply]

A look at the specification tab doesn't help much. I got this...

Product Height 1"

Product Width 1"

Product Weight 2 lbs.

Product Depth 49-7/8'

I see it as just further evidence that the USA MUST metricate. HiLo48 (talk) 00:03, 21 April 2011 (UTC)[reply]

My guess is that they cut off 50-foot lengths of cable, and the process of putting the ends on it uses up 3/4 inch on each end. Why don't they just cut off 50-1/8–foot lengths of cable, you ask? Probably because the spools of cable they get from the cable manufacturer come in even multiples of 50 feet, so if they used slightly more than 50 feet for each cable they'd have an odd length of cable left over at the end. —Bkell (talk) 00:08, 21 April 2011 (UTC)[reply]

By the way, metrication wouldn't make a difference—you'd just have, say, 14.96-meter-long cables instead. —Bkell (talk) 00:13, 21 April 2011 (UTC)[reply]

True HiLo48 (talk) 00:46, 21 April 2011 (UTC)[reply]

What exactly would be the advantage of having the cable length be a round multiple of either feet or meters? It is how long it is. If you want longer, buy longer. If you want shorter, buy shorter. It's not as though it's going to be pulled taut anyway. --Trovatore (talk) 00:31, 21 April 2011 (UTC)[reply]

I see the question more as one about human behaviour. How did such a length come about? Why mention it with such precision? Why not, as many salesmen over many centuries would have done, describe it as 50'?

Though maybe it's worth noting that the other three dimensions above are not so odd. Maybe the 2 lb weight is very precise and is the real defining measurement, with the length being a consequence of the weight. I'd have no idea why that would be the case though. HiLo48 (talk) 00:46, 21 April 2011 (UTC)[reply]

I assume they manufactured whatever is most convenient that is of approximately the correct length, then reported that length to available precision. What is strange about that? --Trovatore (talk) 00:50, 21 April 2011 (UTC)[reply]

Why not describe it as a 50-foot cable? Likely because knowingly selling a cable as a 50-foot cable when it is measurably shorter than 50 feet runs afoul of some product labeling laws. —Bkell (talk) 01:08, 21 April 2011 (UTC)[reply]

That sort of thing seems to happen in the auto industry all the time. Everyone knows that engine capacities are rounded up, when we all know that the manufacturers definitely know a more precise measure. HiLo48 (talk)

They're rounded up for model names. It's not like the manufacturers keep the real capacity a secret; you can easily find it in the owner's manual. --Trovatore (talk) 01:22, 21 April 2011 (UTC)[reply]

Also, as far as I'm aware, automobile engines aren't sold by capacity in the same way that Ethernet cables are sold by length, gasoline is sold by volume, food is sold by weight, and so on. If a cannery produced cans of corn labeled "15 oz" that contained only 14.75 ounces, or a filling station charged you for 10 gallons when you only pumped 9.9, you can bet they'd find themselves in trouble pretty quick. —Bkell (talk) 02:51, 21 April 2011 (UTC)[reply]

The reason for the precision is that, in the US as well as most other countries, consumer protection laws require that products be labeled accurately, or else it's considered false advertising, and fines can ensue. It certainly seems to me that a 1/8th-inch error across a span of 50 feet wouldn't count as false advertising, but maybe somebody in the legal department thought that since the "shortage" compared to 50 feet is consistent, they'd better be as truthful as possible on the packaging. Comet Tuttle (talk) 18:24, 21 April 2011 (UTC)[reply]

The difference is 1/8 foot, or 1½ inches, not just 1/8 inch. —Bkell (talk) 22:25, 21 April 2011 (UTC)[reply]

Aside: I found 49-7/8' very hard to interpret. Is this a standard method? I presume that this is parsed as 49 (- separator meaning "and" or "plus") 7/8 (' feet) more conventionally known as 49.875'. It was the - sign that threw me at first. -- SGBailey (talk)

That's a common convention in the United States when measurements include a whole number and a fraction, though it's more often seen with inches (e.g., 4-3/8″) because non-integer inch measurements are commonly expressed with fractions rather than decimals. (Most rulers and tape measures mark off eighths or sixteenths of an inch rather than tenths.) It's unusual to see a measurement that uses eighths of a foot. —Bkell (talk) 15:27, 21 April 2011 (UTC)[reply]

Just a thought. It could be a typo, intended to read 4' 9-7/8" i.e. 57.875" or 1470 mm. This might be 1.5 m of cable, less whatever is lost in making the connections. AndyTheGrump (talk) 19:42, 21 April 2011 (UTC)[reply]

Thanks for all the responses. There are a few things that could be in play here. Labeling laws, cord length with connectors vs. without, etc. As I said, I was doing a search for Ethernet cable and these are the results. You'll find the 49-7/8' cable near the bottom right above another cable that is labeled as 25-3/8'. Anyone care to explain away the 5/8'? That's more than enough to put connectors on the end. What drew my attention to it so dramatically is the fact that these two "odd" (for lack of a better word) cable lengths are after several other cables whose lengths are even (not partial) number of feet, 6, 14, 50, etc. Dismas|^(talk) 19:58, 21 April 2011 (UTC)[reply]

One further thought. Most metals and I assume plastics expand and contract slightly with temperature. How much is 50 feet of cable likely to vary in length over the usual range of temperatures it's likely to be exposed to? Maybe the length is actually measured at, say, NTP - "Normal Temperature and Pressure" - which stipulates 20C, but for legal reasons has to be quoted at, say, STP - "Standard Temperature and Pressure" stipulating 0C, at which temperature it would be slightly shorter: I remember the difference being mentioned as important in Physics and Chemistry classes, with the added joke that the US favoured higher-temperature standards because many of its labs are in sunny California, while impoverished UK scientists found zero Celsius closer to their everyday reality. It would be reasonable for there to be a minimum standard of accuracy, which length variations for cables shorter than 50 feet might not exceed. {The poster formerly known as 87.81.230.195} 90.197.66.111 (talk) 00:12, 22 April 2011 (UTC)[reply]

I suspect that they tried for exactly 50', but it fell short, possibly because they didn't include the connectors in the calculations, were measuring at the wrong temperature, etc., as mentioned previously. After they determined that it fell short, they either had to revise their processes to increase the length, which could have been expensive, or just update their packaging to indicate the actual length. They chose the later. StuRat (talk) 07:07, 22 April 2011 (UTC)[reply]

Facebook info page[edit]

Hello. The "help" page at Facebook is virtually useless, so I decided to come here and talk to real people. When I set up my Info page, I entered the name of my employer. At the time, the company did not have a Facebook page, so it linked to a blank generic site with the employer name at top. The company has since started a Facebook page, but I can't seem to get the old link to go away. I deleted the employer from my info page, and tried entering it back in as a new employer, but it still takes me to the generic page. How do I make the proper link? Thank you kindly. — Michael J 23:48, 20 April 2011 (UTC)[reply]

I can only think the employer is not deleting properly. I don't mean to tell you how to suck eggs, but I'll go through step by step.

Go to your Profile.
Go to 'Edit profile' (at top).
Go to 'Education and work' (at left).
At the right of your 'Employer' click the cross next to where it says 'Edit' (don't click Edit, click the cross beside it which will delete the employer)
A dialogue box will come up and you must then click 'Confirm'. Your employer field should then be blank.
At this point I would return to your profile; there is a 'View my profile' button just above the employer field to do this, or click the usual 'Profile' link at the top of the page - note that this step should not be necessary, but it's worthwhile to do to confirm that the employer is definitely deleted.
At the top of your profile it should now have an 'Add your current work information' link, where it used to list your generic employer. If so, this is good - click the link and continue below; if not, it hasn't deleted properly and you need to go back and delete it again.
Now back in the 'edit profile' screen, start typing your employer's name again. A list of possibilities should show up. Keep typing the exact name of your employer's page and it should eventually show up in the list. If it doesn't show up, chances are you're making a typo somewhere. You need to go and find your employer's page and check the exact name they're using. They do have to match.
Once you see your actual employer click on the name and it will add to the Employer field. A blue button should then show up that says 'Add job'. You must then click on this to actually add the new job.
If you now return to your profile the proper employer should be listed.

This should work, but it seems with FB there's no guarantees. --jjron (talk) 15:07, 21 April 2011 (UTC)[reply]