Wikipedia:Reference desk/Computing

Welcome to the computing section
of the Wikipedia reference desk.

skip to bottom

Select a section:

• Computing
• Entertainment
• Humanities
• Language
• Mathematics
• Science
• Miscellaneous
• Archives

Want a faster answer?

Main page: Help searching Wikipedia

   

How can I get my question answered?

• Select the section of the desk that best fits the general topic of your question (see the navigation column to the right).
• Post your question to only one section, providing a short header that gives the topic of your question.
• Type '~~~~' (that is, four tilde characters) at the end – this signs and dates your contribution so we know who wrote what and when.
• Don't post personal contact information – it will be removed. Any answers will be provided here.
• Please be as specific as possible, and include all relevant context – the usefulness of answers may depend on the context.
• Note:
  • We don't answer (and may remove) questions that require medical diagnosis or legal advice.
  • We don't answer requests for opinions, predictions or debate.
  • We don't do your homework for you, though we'll help you past the stuck point.
  • We don't conduct original research or provide a free source of ideas, but we'll help you find information you need.

Ready? Ask a new question!

How do I answer a question?

Main page: Wikipedia:Reference desk/Guidelines

• The best answers address the question directly, and back up facts with wikilinks and links to sources. Do not edit others' comments and do not give any medical or legal advice.

See also:

• Teahouse
• Help desk
• Village pump
• Help manual

September 16

'Air' and or 'space' weather simulatiors and designing softwares

What's the best software(s) to create things like this?

Simulation of air flow/fly testing is desirable for bike(s), car(s), plane(s), building(s) and spaceship(s). Can someone help me please? I'm looking for something/a software(s) i.e. extremely good, with which I can stick by as I don't ever wish to worry or change it in my lifetime after learning it, just wish to update it to a newer version... -- Space Ghost (talk) 07:20, 16 September 2015 (UTC)

There are many pieces of software to fit all budgets. Search for "Computational Fluid Dynamics". You will find open source, affordable, and extremely expensive softwares; choose one which suits your needs. Autodesk have CFD software with a free trial. This might work for you. 217.158.236.14 (talk) 09:53, 16 September 2015 (UTC)

Okay. I'm new to this whole thing and thank you for your guidance... Regards. -- Space Ghost (talk) 19:23, 16 September 2015 (UTC)

217.158.236.14: I do possess some Autodesk products and I'm wondering whether I'll be able to draw products on 'CAD', '3ds Max', '3ds Max Design' and or in 'Architecture Revit' then transfer it the 'CFD' software? The thought acquired also because this states about a design study environment... -- Space Ghost (talk) 20:17, 16 September 2015 (UTC)

Why Wikipedia uses paid SSL certificates

Wikipedia uses GlobalSign certificates that costs a lot. Why it doesn't use the free certificates of StartSSL? Hunsu (talk) 09:48, 16 September 2015 (UTC)

It is all about security. You also assume that Wikimedia didn't negotiate a deal with GlobalSign. The lesson is that even on the internet, you get what you pay for. Want an example, Google for GoDaddy nightmares and you'll have a day full of anecdotes to read. Now, try Network Solutions. Hmmm - mostly just user idiocy, not Network Solutions providing poor service. 209.149.113.66 (talk) 12:15, 16 September 2015 (UTC)

StartCom suggests the free certificates offered are very limited and unlikely to be suitable for the Wikimedia Foundation. Also why do you say GlobalSign certificates cost a lot? Their website [1] says may be $2,297 for 3 years for a simple set up. The WMF probably has more complicated needs but even $25k is frankly chump change to the WMF.

P.S. To be clear, I'm not suggesting the WMF does, or should, just waste money. But you have to consider that when you're a large foundation with operating budgets in excess of $50 million, supporting quite a few different domains most with many unique subdomains what may seem like a lot to an individual may not seem quite so much. And other stuff like support, feature set, bundled services (does the WMF get their certificates and domain management from the same company?), support, past experiences etc will likely significantly come in to play. Note also it's likely this sort of stuff is mostly managed by paid staff, probably higher level ones since it's important. So wasting many hours on it isn't cheap, not to mention the negative effect on the people they're trying to reach caused by any problems.

Nil Einne (talk) 14:27, 16 September 2015 (UTC)

Mail list display in online Outlook 365

So my work uses the online version of Outlook 365 as their mailing program. Recently, the line spacing on the mail list (to the left of the reading pane) has gotten much bigger, meaning I can only see 10 or so mails, even on my 23" monitor. I can't seem to change the line spacing in any of the display option, and google is failing me. Help? 131.251.254.154 (talk) 10:00, 16 September 2015 (UTC)

Using it also and have had this problem. Try the following . . .

• On the left, click a particular folder or sub-folder (e.g. Inbox)
• At the top, go to VIEW
• At top left, click Change View
• You're probably currently in the first of 3 choices, "compact"
• Change this to the second choice, "single"

You'll probably have to do this separately for each folder and sub-folder. Hope this works/helps {The poster formerly known as 87.81.230.195} 185.74.232.130 (talk) 13:57, 16 September 2015 (UTC)

Hmm, I don't even have a View tab. Must have been disabled by our friendly neighbourhood sys adminds. Oh well, it was worth a try. Cheers anyway! 131.251.254.154 (talk) 08:17, 17 September 2015 (UTC)

What happens to a regexp when you compile it?

When you compile a regex, what is happening to it? For, example, in Python:

r = re.compile('wiki')
r.match('wikipedia')

instead of:

s = 'wiki'
re.match(s, 'wikipedia')

r and s are somehow different now.

python -m timeit -s 

for both are different. How are r and s different? --Denidi (talk) 15:04, 16 September 2015 (UTC)

When you compile a regex in your code, you can use it over and over with just match. When you use re.match, it will be compiled behind the scenes. So, in both cases it is compiled. It is just a matter of who compiles it - you or the library. In my opinion, it is a matter of taste. It is possible a matter of overhead. In the first example, the compiled r still exists after r.match. In the second example, the compiled regex is trashed after the match. Without getting into the internals of the Python interpreter, it is hard to say if that will end in any real difference. 209.149.113.66 (talk) 15:20, 16 September 2015 (UTC)

A thread on StackOverflow that discusses whether it makes a significant difference. I think the executive summary is: no, unless you're doing thousands of comparisons against the same regexp. The docs say "using re.compile() and saving the resulting regular expression object for reuse is more efficient when the expression will be used several times in a single program"; but also "[t]he compiled versions of the most recent patterns passed to re.match(), re.search() or re.compile() are cached, so programs that use only a few regular expressions at a time needn't worry about compiling regular expressions." AndrewWTaylor (talk) 15:32, 16 September 2015 (UTC)

The above answers provide some good perspective on whether compiling provides performance advantage. Ultimately, performance depends on your use-case and the gory implementation details.

To answer the question directly, "what happens to a regexp when you compile it"... well, let's focus on one specific case, the regular expression engine in Python 2.7.10. You can start by reading the programmer documentation; and you can read the official HOWTO; but those documents just explains the interface, and do not describe the internals of re.compile at all. For that, you need to grab source code to inspect directly.

When you import re, the python environment brings in code from the regular expression wrapper, and from an implementation library (specifically, the one written by "Secret Labs", a small independent software company that distributes free and non-free python software).

How are r and s different? Well, s is a string of characters (represented as a Python data type); and r is a Python object that contains state created from that string of characters. This Python object serves as the implementation of a state machine, which can be applied to other input strings.

We can tear open the sre python code. This code implements the class for the "regular expression object," of which your variable r is an instance. If you read the code, you will find that compile creates a Python object that interprets your regular expression string. It iterates over your input string, looking for a bunch of magic characters - these are the specific sequences of special characters in regular-expression syntax that command the behavior. Finally, it emits "OPCODE"s for the regular expression engine's internal state machine - which is, essentially, a deterministic finite automaton (or something like it).

All the details of performance - including whether the pattern is pre-compiled or re-compiled each time you use it - are defined in the "cache" inside Python's regular expression engine.

When you run a "match" or any other of the standard regular expression utilities on a string, the regular expression engine will either re-compile a new regular expression tokenizer object, or use one from the internal cache. Then, it will call into the function that represents the state-machine for that regular expression, which will iterate over your input string, and produce an appropriate output.

All told, the code for regular expressions in CPython-2.7 is written in pure python and is only about three thousand lines long. It would probably take a few hours to really read and understand the general flow, and it might take weeks or months to master the details; but that's what is happening. In other regular expression engines - like the ones in your standard Java or perl distributions, there's a similar process.

If you would like to read what Larry Wall has to say on regular expressions, He is regarded by many as the Creator (although there are many doubters, and there are many students of an older magic). Larry Wall has spoken by way of several Apocalypses. This is perl programmer humor, but it's literally true. You may wish to consult Apocalypse 5, Pattern Matching, which is from year 2002 and is very obsolete; but is a great overview: regular expression culture is a mess. There exist many variants, many poor implementations, many contradictory standards; and the result is a widespread and terrifying set of use-cases; so many pieces of code where a regular expression engine is abused beyond reason; so many instances where a piece of code only works in certain carefully-balanced conditions.

Nimur (talk) 16:40, 16 September 2015 (UTC)

In short, regexes are compiled into bytecode. The bytecode interpreter is written in C (_sre.c). The bytecode does not define a deterministic finite automaton (finite automata can only match regular languages, while Perl-style "regexes" can match much more than that despite the misleading name). It is nondeterministic (evaluated using backtracking, similar to Prolog) and it has state beyond the program counter, such as the contents of previously matched groups for \# backreferences. -- BenRG (talk) 20:55, 16 September 2015 (UTC)

I may be missing something, but in one case isn't the RE compiled at compile time whilst in the other it is compiled at run time? (I know nothing about python and a little about perl.) -- SGBailey (talk) 23:42, 20 September 2015 (UTC)

No, Python (well, Cpython, which is what people mean when they say Python without qualification) isn't compiled (well, it is, but effectively at "run time"). The difference between the two ways above is whether the result of the regexp->bytecode compile is stored explicitly in a variable r, or cached implicitly by the regexp library (so, for most cases, no worthwhile difference at all) - that's more of a hint to other programmers (hey, we're using this regexp again later) than to the system. Tangentially, the only place I've heard of where a regexp is really really compiled is the .NET implementation, which has the option to compile not to a regexp-specific microVM like the one BenRG talks about, but instead to .NET CLR bytecodes, which the .NET runtime JIT can then compile down to actual machine instructions. -- Finlay McWalterᚠTalk 00:09, 21 September 2015 (UTC)

ASP.NET authentication question

I have run into a problem at work. This is about authenticating a user into an ASP.NET C# web application using username/password Forms authentication.

The thing is, the username/password that the user types in and the application verifies against the database needs to be different from the username that the application actually authenticates and stores in the authentication cookie.

I am using an ASP.NET Login control to do the username/password authentication. The username/password verification is done with a custom-made MembershipProvider class. Is it somehow possible to change the authenticated username after a successful verification in the OnLoggingIn, OnAuthenticate or OnLoggedIn events? JIP | Talk 19:42, 16 September 2015 (UTC)

September 17

Shanks' square forms factorization

In Shanks' square forms factorization, in the middle of the algorithm is

until $Q_{i}$ is a perfect square at some even $i$.

Initialize $b_{0}=\left\lfloor {\frac {\lfloor {\sqrt {kN}}\rfloor -P_{i-1}}{\sqrt {Q_{i}}}}\right\rfloor ,P_{0}=b_{0}{\sqrt {Q_{i}}}+P_{i-1}$

In the calculation of $b_{0}$ and $P_{0}$ it refers to $P_{i-1}$. The $P_{i-1}$ actually refers to the last value of P calculated in the previous section. Can someone fix this? Bubba73 ^{You talkin' to me?} 01:49, 17 September 2015 (UTC)

Looking for an Ethernet Splitter that lets both cables deliver at the same time?

So I've got an ethernet cable sticking out of my wall and I want to jam something in it so that I can connect two ethernet cables to it, and then me and my roommate can both have wired internet at the same time. I bought a splitter off of Amazon but it turns out it was toggle only; couldn't have both connections running at once. So I returned it and bought another; well if that one didn't have the same problem! Now I'm paranoid; can anyone recommend one that actually lets 'em both work at the same time? 2605:6000:EA01:7E00:BDDD:DDAA:F59F:6C9E (talk) 06:26, 17 September 2015 (UTC)

You want a switch or hub. As the articles say, hubs used to be preferred for that kind of application because of price, but switches are now about the same price, and generally preferred. A quick search for "4 port switch" gives a number of options in the £5-£20 range, depending on whether you want gigabit. Why not a 3 port switch? Well, because the only result that came up from a quick google was about ten times the price of a 4 or 5 port.MChesterMC (talk) 08:28, 17 September 2015 (UTC)

You need a four-port switch/router. Netgear make some simple ones that will work. — Preceding unsigned comment added by 217.158.236.14 (talk) 08:23, 17 September 2015 (UTC)

Thanks you two! I'll buy one of those! 2605:6000:EA01:7E00:BDDD:DDAA:F59F:6C9E (talk) 10:44, 17 September 2015 (UTC)

Which Javascript library is 'tkjs-ni/sources/NI-core'?

I have some Javascript code that calls the following libarary:

    var NI = require('tkjs-ni/sources/NI-core');

What is it? Googling "tkjs-ni" or "NI-core" got me nothing useful. My other car is a cadr (talk) 08:01, 17 September 2015 (UTC)

That is a relative URL. Without the domain name, it is useless. Nobody is going to go to every website in the world to see if http://every.website.possible.placed.here/tkjs-ni/sources/NI-core exists. 209.149.113.66 (talk) 13:40, 17 September 2015 (UTC)

It's not a URL. I explicitly asked about Javascript libraries, and require(...) is how modules are loaded in nodejs, so obviously 'tkjs-ni/sources/NI-core' would be the name of package. If you don't know the answer then please don't waste Wikimedia Foundation's bandwidth with your ignorance.My other car is a cadr (talk) 07:13, 19 September 2015 (UTC)

Just as a wild guess, I checked to see if National Instruments has a Tcl/Tk javascript package. They do. It is possible that they distribute it as tkjs-ni (Tk javascript from National Instruments) / source (the source code) / NI-core (National Instruments core module). However, that is just a possibility. The code from your website could be something completely different. Without the website that it came from, it is impossible to know. 209.149.113.66 (talk) 14:25, 17 September 2015 (UTC)

Levenshtein distance

The recursive algorithm is referred to as "straightforward", yet I don't understand it (and the recursive definition of Levenshtein distance that it is based upon) ):. Can someone please elaborate on the relation between the definition "Levenshtein distance between two words is the minimum number of single-character edits (i.e. insertions, deletions or substitutions) required to change one word into the other" and the recursive definition & pseudocode? thanks! — Preceding unsigned comment added by 212.179.21.194 (talk) 10:31, 17 September 2015 (UTC)

The recursive solution begins with a stop case. If either string is empty, it returns the length of the other string. That is Levenshtein distance. The distance between "" (empty string) and "ABCDE" is 5 - the length of "ABCDE". Now, if we didn't hit the stop case, we look at the last character to see if the same. NOTE: We could go from the first character to the last character or the last character to the first character. It doesn't matter. The distance between "STAIR" and "HAIR" is the same as the distance between "RAITS" and "RIAH". So, if the last characters of the strings are the same, the distance is not increased. If they are different, there had to be at least one change, so 1 is the cost for this character position. Next, three recursive calculations are made. One is for inserting a character. One is for deleting a character. One is for substituting a character. All of these take place before the last character. First, what if we deleted a character from the first string? If the strings were "STAIR" and "HAIR", it gets the cost of "STAI" and "HAIR" and adds 1 (for the delete operation). Then, it checks delete from the other string: "STAIR" to "HAI" plus 1 (for that delete). Then, it checks the rest of the strings, using the cost calculated earlier by comparing the last character. Whichever one of those calculations is the least (more than 1 can be least) is returned. 209.149.113.66 (talk) 13:38, 17 September 2015 (UTC)

Thanks you very much for the answer! there is one thing I still don't get: initally you say that "three recursive calculations are made. One is for inserting a character. One is for deleting a character. One is for substituting a character." but after that you mention two deletions and no insertion: " First, what if we deleted a character from the first string...Then, it checks delete from the other string.... Then, it checks the rest of the strings, using the cost calculated earlier by comparing the last character". 109.160.236.195 (talk) 16:57, 17 September 2015 (UTC)

The issue here is that insertion and deletion are the same operation. It is just a point of view. If I convert "EAT" to "BEAT", I inserted a B. I could also say that I deleted a B if I go from "BEAT" to "EAT". Since Levenshtein distance gives the same result both ways, we can say "insert" and "delete" are the same thing. That is why the first two recursive calls are for "insert/delete", but I just said "delete" when I got to the bottom of the description. The only tricky one is substitute. That is technically a delete/insert combination, but we count it as one. 209.149.113.66 (talk) 17:04, 17 September 2015 (UTC)

Thanks again, this explanation may be worth adding to the article. 109.160.236.195 (talk) 19:05, 17 September 2015 (UTC)

Open source software about Environmental management system

Hi, I'm looking for an open source software for Environmental management systems to manage for example the ISO 14001 standard. All I found online are expensive softwares without even a shareware version. --Helmoony (talk) 14:34, 17 September 2015 (UTC)

What charger should I buy for this?

I just asked a question about ethernet splitters, and I found what they told me to buy in my attic. However, it's missing its power cable. the little words under the power port says "7.5VDC. 0.8A". Is that a 7.5 Volt, .8 Amp cable? Are those common? Online I can only find 7.5Volt, 2-3Amp chargers closest.2605:6000:EA01:7E00:E921:DEE2:3B6:72AD (talk) 19:18, 17 September 2015 (UTC)

That is the V and A for the power adapter. It is likely a circular plug also. You need to know the polarity. There should be a circle with a dot in the middle. Is the dot in the middle - or +? With that, you can get a universal power adapter (relatively cheap). Find a plug that fits in the hole (it will come with 5 to 15 plugs for various sizes), set the polarity, set the voltage. The universal ones tend to provide more than enough amps for most devices, so I wouldn't worry about that. 209.149.113.66 (talk) 19:39, 17 September 2015 (UTC)

There should be a polarity symbol sign beside the voltage/ampage, or on the base of the unit. Failing that, if you do a websearch for the model number, you might find it. Centre-positive is now ubiqutious, but centre-negative was in a sizeable minority 10 years ago. You should also check the maximum speed of the hub/switch; MS-Windows should tell you in the Network icon on the systray. If the hub is old enough it might be 10Mbit/sec only, in which case it will slow down your connection, and should be replaced. LongHairedFop (talk) 20:05, 17 September 2015 (UTC)

The core of your question seems to be asking if the 7.5VDC 2 or 3 amp will suit a device that says 0.8A. The answer is YES, the value on the device is just the 'minimum value that you should use. In most cases you would just want to use the "matched" transformer because using a larger one will cost more, sometimes be less efficient, weigh more, etc... but if you have no other choice, or if 0.8 is unusual and expensive but 2 amp one is cheap and easy to get, just get the cheap one. The thing you should NEVER do, is use a 0.8Amp charger for something that says it require 2Amps. Vespine (talk) 23:07, 17 September 2015 (UTC)

September 18

Help identifying printer

I remember seeing a video a few years ago of a computer printer that used a normal ballpoint pen as its ink source. It worked much like a seismometer with an arm that held the pen and moved it from side to side, pressing down as appropriate to render the printout. Except instead of drawing earthquake graphs it could print text and monochrome images from a computer. Does anyone know what this type of printer is called? Is there a Wikipedia article on it? Thanks for your help — Preceding unsigned comment added by 117.178.158.168 (talk) 14:58, 18 September 2015 (UTC)

That sounds like a pen plotter. -- Finlay McWalterᚠTalk 15:07, 18 September 2015 (UTC)

... although a pen plotter didn't quite use a "normal ballpoint pen", but rather a technical pen system with a short, stubby pen made to be held by (or screwed into) the plotter's head: they're not ergonomically suitable for using hand held. And, like other technical pens, both the pens and the inks were quite expensive. -- Finlay McWalterᚠTalk 15:14, 18 September 2015 (UTC)

Thanks that seems to be exactly it! The article even says "Ball-point plotter pens with refillable clear plastic ink reservoirs are available. They do not have the fading or wear effects of fiber pens, but are generally more expensive and uncommon. Also, conventional ball-point pens can be modified to work in most pen plotters." Now that I have the name I even found some holding a normal pens [2] [3]. Maybe the one I saw was a diy project or something. Anyway, you answered the question thanks! — Preceding unsigned comment added by 117.178.158.168 (talk) 15:31, 18 September 2015 (UTC)

Of peripheral interest, there were a few low cost plotters of this type produced for 8-bit computers of the 1980s era, e.g. the Atari 1020. --LarryMac | Talk 21:01, 21 September 2015 (UTC)

Free LAMP stack hosting

I will be teaching a course in LAMP stack development. I'm trying to find a free LAMP stack hosting service for the students to use. I assume that any service I find will have ads in some way, but I haven't been able to find any free LAMP hosting. Everything that I find claims to be free and, just when you get to the point of setting up the server, you have to choose a monthly billing plan. Are there still absolutely free LAMP stack hosts available? I don't care if you can't assign a domain name to the website. I just want to be able to get to the CLI and do development. 209.149.113.66 (talk) 15:13, 18 September 2015 (UTC)

"Free" Web hosts are notoriously terrible. Most won't give you CLI access, and the ones that do will have a strictly limited environment. You generally won't be allowed to install things, modify many settings, etc. If you want a decent hosted development environment you'll have to pay at least a little. Are you sure the institution you're teaching for won't cover the cost? Many large institutions with their own IT infrastructure would host it themselves, but of course I know nothing about your situation. --71.119.131.184 (talk) 18:22, 18 September 2015 (UTC)

If this is only for your students, why not get a few old and obsolescent computer and create your own hosting system? Just create enough VM's (or virtual severs) for everyone. The bandwidth is bound to be small (unless your students are also into porn). --Aspro (talk) 19:22, 18 September 2015 (UTC)

I take it that that the OP is from a third world county where most people starve and money is in very short supply. If he lives (in say) Africa he may be able to join WebAfrica Their Linux Premium Plus is less than I spend on a single night out. If you would like to email me, I will Pay Pal transfer a years hosting for you and your students. Students are our future. --Aspro (talk) 21:21, 18 September 2015 (UTC)

Thank you, but this isn't a third-world country course. It is more like a third-world county. I will be teaching at a community college with very few resources. The students are all on tuition assistance from the state. They have a computer lab, but it is windows-only and doesn't provide any user space. There is only the "guest" account that gets wiped with a new image every day. So, my plan is to have the kids put Putty on a USB stick and run it to connect to a LAMP server. Then, do all development there. I don't personally own a LAMP server that they can use, so I am trying to find one that they can use for the 4 months of the semester. 209.149.113.66 (talk) —Preceding undated comment added 12:43, 21 September 2015 (UTC)

Could you not install one in a VM on one of your Windows hosts? (Or all of your windows hosts - that way each student could have their own!).--Phil Holmes (talk) 14:44, 21 September 2015 (UTC)

Wow, I didn’t know things were that bad in South Carolina. Your in one of the riches countries on this planet but your students, out there in the sticks, are not sharing in the wealth (Barack are you hearing this). What pops into mind, is that although you are a Windows only environment, leave the PuTTY to one side for a moment. A lot of windows only environments now use Linux servers for ease of maintenance. What I think may be a solution is to use a Linux Server on a pen-drive that has a windows compatible NAS file, (to keep life simple). Ubuntu Servers and and many other Linux distributions will go onto a bootable pendrive. Here are some others: Portable Web Servers Category. Remember, pen-drives can wear-out so always back up and have a few spare servers. I can't add any more to that because I don't know your exact circumstances are. Also, as you live in one of the richest countries on this planet I don't feel like subsidizing your badly funded education (?) system without your Government sending me a fair remuneration (Barack are you hearing this). No offence meant your students, I wish them well and sure they will do well under your mentorship. As -Phil Holmes suggests. I think you have a very good reason to ague that you need a VM that doesn't get wiped every day. What it the IT managers job for if it not to provied IT support? Get back in the future and let us know how you get on.--Aspro (talk) 15:08, 21 September 2015 (UTC)

Final thoughts: Remember, setting up a VM is is not going to cost your faculty any money really. The are still paying their IT manager regardless of whether he is sting down eating doughnut as he surfs the web or picking his nose. All they need to do is get him to pull his finger out and provide you and your students with what is needed for the purpose of education. Still, I like the pen-drive solution though. Not as good a proper solution but it will give your students the experience of how to carry around their very own server. Maybe introduce them to both. They can then play, experiment and learn with adim access on the pen-drive and mess it up all they want -with no harm done. Just a thought.--Aspro (talk) 15:40, 21 September 2015 (UTC)

I am trying what, for me, will hopefully be the easiest route. I'm trying to get the IT staff to allow me to put a LAMP server on the school's network. I will have complete control over the installation and configuration and I can create an account for each of the students. The hard part is convincing the IT staff that it won't be a security risk. They are already freaking out and labeling me a major security threat for suggesting that the students plug a USB drive into their computers. 209.149.113.66 (talk) 13:17, 22 September 2015 (UTC)

DOWNLOAD AUDIO FROM VIDEO

How can I download only the audio from an youtube or any other online video? Thank you.175.157.53.223 (talk) 15:23, 18 September 2015 (UTC)

This online service [4] might be helpful. SemanticMantis (talk) 19:04, 18 September 2015 (UTC)

Also, I think DownloadHelper (which is free) lets you choose the format to save the file in, and some of the formats are audio only. OldTimeNESter (talk) 12:29, 19 September 2015 (UTC)

Creepy Facebook friend suggestions -- Facebook Google connection?

I'm getting increasingly concerned about where Facebook gets its "People you may know" friend suggestion list from. I'm not very active on Facebook, and have few Facebook friends. Yet several of the suggestions are very precise. Problem is, I have no idea where they get the information from, and I don't like the idea of having all sorts of details about my personal relationships floating around in cyberspace. Here is the list:

1. Wife of nephew. 11 mutual friends. OK.
2. Husband of colleague. No mutual friends. I am not a Facebook friend of this colleague. NOT OK.
3. Niece, 8 mutual friends. OK.
4. GF of son, 10 mutual friends OK.
5. IT consultant used by my employer. Extensive communication a couple of years ago, including using my private Gmail account. No relationship through friends of friends as far as I know. Definitely NOT OK.
6. GF of nephew, 8 mutual friends, OK.
7. Nephew, 9 mutual friends, OK.
8. Relative, 7 mutual friends, OK.
9. Close colleague, no mutual friends, but Gmail contact. NOT OK.
10. Close colleague, no mutual friends, but Gmail contact. NOK OK.

I have not fallen into the "please give us your email password" trap. So what's going on?

--NorwegianBlue ^talk 21:46, 18 September 2015 (UTC)

Even if you did not share your Gmail contacts, there people obviously could have shared theirs (which includes you). It's not enough that you take care about not releasing too much private information. People around you should also take care.

Add to it that FB can analyze the information it gets. FB could also have gotten the information through their own mining efforts. You and your colleagues disclosed your place of work somewhere, they picked the information and concluded that you could know each other. They obviously use methods more refined than just looking at the friends of friends as potential FB buddies. It's all statistics, but probably it does not have to be a very sophisticated algorithm. --Scicurious (talk) 22:09, 18 September 2015 (UTC)

WP:OR - I share your concern. FB has suggested many people that I may know with zero mutual friends that are concerningly accurate. One unsubstantiated idea - either you have the person's phone number and name in your phone contacts, or vice versa. At least, I think this has been a possible explanation for my experiences. I can't think of any other way that they would know that I know past landlord - we live in different states, have zero personal connections or friends in common etc. I can't think of any way this would have come up through traditional statistical inference. But he likely has my real name and phone number on his phone, and I do have that phone number in my FB profile... of course this wouldn't make sense if neither party used the FB app on a smartphone. SemanticMantis (talk) 22:25, 18 September 2015 (UTC)

September 19

Automatic data sent while using internet

How do I check 1) what datas are/were sent/getting sent and 2) How do I stop it? -- Space Ghost (talk) 03:36, 19 September 2015 (UTC)

[1] Use an open source operating system such as Linux and use tcpdump. You can try using the same utility on Windows (it's called WinDump) but you can never be sure that Windows or some malware isn't hiding the traffic from your monitoring tool.

[2] Use Tails (operating system). It boots from a CD/DVD and doesn't touch your existing operating system.

--Guy Macon (talk) 04:36, 19 September 2015 (UTC)

Thanks Guy. Option 2 sounds familiar - SementicMantis and Aspro spoke to me about a 'Live CD' option for Linux OS once, also gave me link(s) to download which I've bookmarked, I did not come around to download it yet due to internet usage problem... -- Space Ghost (talk) 18:45, 19 September 2015 (UTC)

I would suggest it makes no sense to trust monitoring on the OS period, particularly when there are far better alternatives. Particularly since if you are asking this question, it's unlikely you have the ability to analyse the source code of either tcpdump, or the network stack of the OS you're using. Since the traffic has to be going somewhere, unless it's a mobile device using a cellular netowkr, you probably have control, or can have control over the next hop. So it makes far more sense to monitor the AP or router the device is connected to. You still have to trust the device you are using for monitoring and the tool on it, but in that case it's a specialised system so far more likely to have been properly look in to plus the kind of bugs that would cause traffic to be missed are different and probably less likely to occur. (In one case you have be sure that the monitoring tool is probably attached to the network stack and no traffic can be sent without being captured. In the other case you just have to be sure everything received is captured.) Using a tool on the sending side may still be useful, since you then can more easily determine what's sending the data etc, but if you're talking about trust you should generally have something on the other side to be sure. Nil Einne (talk) 06:18, 19 September 2015 (UTC)

wireshark and netstat to view, peerblock to stop. — Preceding unsigned comment added by 2A03:2880:3010:BFF9:FACE:B00C:0:1 (talk) 18:32, 19 September 2015 (UTC)

Awesome! Thanks! -- Space Ghost (talk) 18:45, 19 September 2015 (UTC)

Ubuntu will automatically send your desktop search terms over the Internet to a large number of third parties unless you configure it not to (see e.g. [5]). All major Linux distros can track everything you install through their standard package repository, in contrast to the traditional Windows model where you get most software from distributors not affiliated with Microsoft and they generally have independent update mechanisms. Microsoft is apparently starting to collect installed-software telemetry too, and they now have a Windows Store over which they exercise dictatorial control, but they are only able to get away with that now because it was pioneered and made acceptable by other companies that people don't hold to the same standards as Microsoft.

Independently of what OS you use, ad networks can track your browsing behavior across all pages that display their ads, social networks can track you across all pages that have "Like" buttons, and other third-party services like Disqus can do the same, via third-party cookies. Ghostery can help with this.

Many software products, including browsers, have a "help me improve this product by collecting anonymous usage statistics" option that may be turned on by default. At least Chrome, Firefox and Safari use Google Safe Browsing. There may be other default-on features of this type so you generally have to pore over all the settings and figure out what you want to turn off.

In short, the idea that using Linux is going to protect you from monitoring is nonsense. The general social trend toward acceptability of this sort of monitoring affects all OSes and applications. Tails is a different matter since it's specifically designed and configured to avoid this stuff, but it's also a very limited way to use a computer. -- BenRG (talk) 21:28, 19 September 2015 (UTC)

Hello, sorry! I forgot about this post, I was busy with some other thoughts...with daydreams/OBEs. I understand. I'm using google chrome, I'm logged into the browser application with my google e-mail ID.

'Ubantu' software is available where I am, along with 'fedora' and 'Linux Red Hat'. I've been dying to get into UNIX OS for such a long time, finally I found a 'Live CD' method, still it feels like miles away, in order to get into it...cause I don't know nothing about it; I'll definitely need you guys to help me during the infancy and toddling period I will try to get into it the day when I become free, that's if I don't start daydreaming/having OBEs...

Thanks friends. I appreciate it. I'm gonna read into it on the upcoming festival day. Regards. -- Space Ghost (talk) 18:21, 20 September 2015 (UTC)

Space Ghost, again, when you choose which live CD to use. I highly recommend Tails (operating system). All you do is boot it up, wait a bit while it does its magic, and suddenly you are using Firefox through a secure connection on an operating system that is specifically designed for the one task of not revealing anything about your identity, location, or IP address. I trust it to protect industrial secrets when I am in China working on toy designs through a WiFi connection owned by a company that badly wants to see what I am doing on the internet. --Guy Macon (talk) 01:10, 21 September 2015 (UTC)

Thank you very very much! I really really need something like this… -- Space Ghost (talk) 06:18, 21 September 2015 (UTC)

How exactly does this help? The OP is asking about how to check what data is getting sent and how to stop it. It's extremely unlikely that this will even answer the first question, nor the second. A Quest For Knowledge (talk) 01:58, 21 September 2015 (UTC)

Are you sure you are posting to the right thread? Because in this thread the OP was quite specific:

"I've been dying to get into UNIX OS for such a long time, finally I found a 'Live CD' method, still it feels like miles away, in order to get into it...cause I don't know nothing about it; I'll definitely need you guys to help me during the infancy and toddling period"

So we know he is concerned about "Automatic data sent while using internet". We know he wants to get into Linux, and we know that he wants to try using a live CD. Put those together and you get Tails. Plus it is the Live CD I use the most, and I am going to make a point of helping him "during the infancy and toddling period". --Guy Macon (talk) 21:16, 21 September 2015 (UTC)

What OS are you using? Are you using Windows, Android, iOS, etc.? If you're using Windows, and are just concerned about web browser data, you can just open developer tools and look at the data. If you are using Windows and want to see all data regardless of app/browser, Fiddler or Wireshark will work. If you're using iOS, I don't think it's possible (unless perhaps you want to rootkit your phone). I don't know too much about Android, so I can't say one way or another. A Quest For Knowledge (talk) 01:44, 21 September 2015 (UTC)

PC – Windows 7 Ultimate 32-bit OS. Data viewing is not as important as ‘data sent’ blocking while using the internet; it’ll be good to know though. 1stly, I’m on ‘pay bite as you go’, it hurts my feelings when my eyes suddenly catch the 'Data sent' information on the modem’s window, what shouldn't be more than ‘Data received’ information... -- Space Ghost (talk) 11:33, 21 September 2015 (UTC)

OK, I am going to get a bit technical here, so ask questions if I am not clear. Read these articles:

• http://arstechnica.com/information-technology/2015/08/microsoft-accused-of-adding-spy-features-to-windows-7-8/
• http://www.extremetech.com/computing/213183-once-more-with-tracking-microsoft-backports-privacy-invading-windows-10-features-to-windows-7-8
• http://www.ghacks.net/2015/08/28/microsoft-intensifies-data-collection-on-windows-7-and-8-systems/

Basically, as part of the Windows 10 rollout, Windows 7 started "phoning home" and sending data to Microsoft. The articles above tell you how to get it to send a lot less. And of course, as we discussed before using a Live CD like Tails means that no data is sent without you personally sending it. I personally would never access, say, my bank's website from Windows. --Guy Macon (talk) 21:16, 21 September 2015 (UTC)

September 20

Open source calculator that looks like a real Texas Instruments calculator

What software calculator has the closest look-and-feel to a real hand calculator like a TI-83 or the TI-84? For a picture look at the TI-83 series. I know that there are better alternatives to perform calculations using a laptop. Sagemath is an example, among others. However, I am interested in getting used to the calculator allowed at many examination, not at the calculation result properly. --Scicurious (talk) 00:00, 20 September 2015 (UTC)

I have heard good things about Wabbitemu [ https://wabbit.codeplex.com/ ]. According tto this PDF page, it runs a downloaded image of the code that runs on the actual TI calculator, so everything should act the same, including any bugs (you don't want to discover that your calculator has a bug that your emulator doesn't while taking an important test!) A web search also turns up TilEm [ http://lpg.ticalc.org/prj_tilem/ ] and jsTIfied [ https://www.cemetech.net/projects/jstified/ ]. I tried Virtual TI a long time ago and was not impressed, but maybe there is a newer version that is better.

Please come back and update this with the results of your experiences with the above emulators so the next person can benefit from your experience. Thanks! --Guy Macon (talk) 01:36, 20 September 2015 (UTC)

Windows 10 - msconfig

In Windows 10, msconfig, General tab, I select the radio button for "normal startup", but it won't save it (even if I Apply or OK). It always goes back to "Selective startup", with a square in "Load system services", a check in "Load startup items", and a check in "use original boot configuration", the latter of which is disabled. Is this normal? Is it OK? Bubba73 ^{You talkin' to me?} 05:22, 20 September 2015 (UTC)

I just tried it and I'm experiencing the same behavior. Sounds like a bug. Do these instructions help? www.ehow.com/how_7390089_repair-msconfig.html A Quest For Knowledge (talk) 14:59, 20 September 2015 (UTC)

For some strange reason, ehow.com is blacklisted, so I had to remove the http:// from the URL. A Quest For Knowledge (talk) 15:00, 20 September 2015 (UTC)

eHow is a low quality content farm which outsources writing to anyone and rewards those authors on the traffic they pull in - see MediaWiki talk:Spam-whitelist/Archives/2010/08#ehow.com. -- Finlay McWalterᚠTalk 15:33, 20 September 2015 (UTC)

Sure, if you're using it as a reliable source in article space. The filter should be smart enough to distinguish between article space and the Reference Desk. A Quest For Knowledge (talk) 15:53, 20 September 2015 (UTC)

You can request a whitelisting for the RD if you want. I myself have occasionally used eHow, however it's something I avoid for reasons besides the blacklist. Anyway I think you're missing the point of the blacklisting which is to discourage spammers, no matter what part of RD they target. Whether spammers are likely to target the RD is hard to say. Nil Einne (talk) 16:21, 20 September 2015 (UTC)

Ammm, two humans I know who returned to 'Windows 7' because 'Windows 10' is not compatable with certain things. -- Space Ghost (talk) 18:26, 20 September 2015 (UTC)

Those instructions don't seem to address this problem (except I'm not sure if it is a problem, I haven't seen any ill effects). Bubba73 ^{You talkin' to me?} 05:22, 21 September 2015 (UTC)

Am I making a hash of this idea?

This could be under Computing or Math. Computing wins (unless y'all don't agree).

I'm looking to host a collaborative project, where each user can choose to remain anonymous. However, I'd like to have a system whereby those who chose the anonymous route can later change their mind.

Let's take Wikipedia as an example. Anyone can anonymously add something or make a change, and there's no reasonable way to find out who they are.

If, in the future, an editor, for reasons of their own, decided to claim credit for a particular edit, there's no way for that editor to prove that they're the one who made that edit, if they were originally anonymous.

I've heard of trapdoor functions. I'm wondering if it's possible and/or reasonable to use these for this project.

My idea is that each time the user contributes, they include a whichamacallit (hash?), which is the product of two numbers. They keep those two numbers secret until such time as they wish to let people know that they were the contributor. My (limited) understanding of the trapdoors is that if they know the two numbers, that's almost incontrovertible proof that they are the perpetrator.

Is this idea feasible? Not having a lot of computer smarts, or knowing a whole lot of math, how would I implement this, without investing a lot of (or any) time or money for what I'm sure has been done.

Please keep in mind that I'm NOT suggesting that this is a good idea for Wikipedia. That's just an example that I feel more than a few readers are passingly familiar with.

Also, there's no need for super-security for my intended use. I'm not worried about a super-power breaking through the users identity.

Thanks for any help or ideas you might come up with. Bunthorne (talk) 21:40, 20 September 2015 (UTC)

Yes, user should submit a hash of their identity and contact information.

For example, in Linux I can type

khan@ToshiK:~$ echo 3dcaddy | md5sum
6a8d1f3d555c1ab285ec66854d15c4eb  -

Reverting 6a8d1f3d555c1ab285ec66854d15c4eb to 3dcaddy is quite difficult.--3dcaddy (talk) 21:57, 20 September 2015 (UTC)

@3dcaddy: sorry to interject like this
echo adds a newline by default, so your example isn't hashing the 7 ASCII character sequence "3dcaddy" but the 8 ASCII character sequence "3dcaddy\n". Use echo -n to suppress the newline, which yields the correct md5sum 5d1c09823512e0aee5080c5858968945 -- Finlay McWalterᚠTalk 14:50, 21 September 2015 (UTC)

Yes, that's right. The OP could also put everything in a file, and run "md5sum file".

Thanks for the quick response. I didn't realize it could be that easy. Unfortunatly, I don't do Linux (although I realize I should). Is there an easy way to do it with windoze? Bunthorne (talk) 22:06, 20 September 2015 (UTC)

Never mind. I did a quick search on "md5", and found a bunch of answers. Thanks for the tip. Bunthorne (talk) 00:32, 21 September 2015 (UTC)

I had marked this "resolved", but removed it since some of the responses brought up other questions. See below Bunthorne (talk) 20:23, 21 September 2015 (UTC)

One thing to bear in mind is that the high speed GPUs (edit: also ASICs, although only people with a specific goal in mind tend to have these) can compute certain hashes like MD5, combined with their popular usage to store passwords in some circumstances means that coming up with the reverse for a short string hash can actually be fairly trivial, even just an internet search. This didn't work with 3dcaddy (if we ignore this page), but would have worked with "bebopjazz1" [6] (random example I came up with when searching for something) or "BillGates" [7] and yes, even Bunthorne [8] [9].

If you get people to use a salt, you'd eliminate rainbow tables, unless people get interested enough in your website to develop a rainbow table for it (as I assume you'd need to use a fixed edit: known salt), but still won't eliminate the GPU problem.

If we assume most people are only going to be hashing their name or some pseudonym and most of these don't have numbers, and so only use upper and lower case letters and perhaps space and underscore, I suspect up to 11 characters may be feasible for someone without that much more resources than a dedicated gamer. ([10] is a little old and doesn't have exactly what you're looking for but is close.) It gets even worse if we decide it's most likely a real name from the Anglophone world or something of that sort. (Similarly if you're only using numbers, probably something like 18 decimal digits could be feasible.)

You'd generally want people to include at least 20 characters IMO and make sure it's more than just a name, maybe 30 or even 50 to be safe considering possible future advancements. An alternative (edit: or perhaps combined with a decent sized string) is to use a hash designed to be slower on GPUs edit: (and hopefully ASICs) like bcrypt or scrypt, although I'm not sure how well cryptologically tested these are.

Also due to the nature of hashes, people will have to record or remember exactly what their original string was including cases etc. Okay you could make a brute forcer for them, if they have a good idea of what it is, but that would complicate things somewhat. (Sufficient info would also reduce possible disputes if person A claims the hash represents them and person B who were the first to say it was them must have somehow reversed it.)

Nil Einne (talk) 07:38, 21 September 2015 (UTC)

Just as a historical note: in science there is always a conflict between wanting to publish an imporant new result in order to ensure you are recognized as being the one who discovered it, and wanting to investigate further in order to produce a more complete report. These days, as I understand it, scientists will first publish a short letter summarizing their finding and follow it later with a full paper. But centuries ago, what they did first in this situation was to publish an anagram of their result, and then when they were ready to do a full report, they would provide the explanation to the anagram. See anagram#Establishment of priority for examples. Of course if they were being tricky it was always possible that the same letters would anagram to more than one relevant sentence, saying different things! --174.88.134.156 (talk) 03:39, 21 September 2015 (UTC)

In the modern publish or perish environment of academics, it is very common to turn everything into multiple publications. For example, I had an idea that I thought was obvious. Everyone I spoke to said it was not obvious and apparently wrong. So, I implemented it and demonstrated that it worked. Then, I published a summary of the background information (summary papers are popular for textbook chapters). Then, I presented/published the general idea at a conference. Then, I published my findings. All together, the time between the summary paper and final publication was three years. So, I had three years to publish final findings while still demonstrating the work with the initial publication as a "guess what you can do with this knowledge" part of the summary paper. As an aside - I also take part in the "publish every paper as many times as possible" method of publication. For example, a paper on track management of hard drives became a paper on track management of radars, which became a paper on track management of cell towers. I'm now working on a paper for track management for a multi-signal bus in a cell processor system. 209.149.113.66 (talk) 11:50, 21 September 2015 (UTC)

It depends a little on how anonymous you want the edits to be. Using a hash of a secret (as proposed above) is fine if you want to know that one person made a set of edits, but not know who the person is. If you want stronger anonymity, where you don't know who made each edit, and can't tell which edits are made by the same person, then you need to include something unique into the hash. If this unique element is public, then it's still possible for the person to verify that they made that edit. For example, the content of the edit, or the time at which the edit was made could be added to the secret and then hashes (e.g. if I choose a secret of "Dave1", then a hash for this edit could be generated from the string "Dave1-10:40, 21 September 2015 (UTC)", and there would be no way of linking this to my next edit without knowing the secret). MChesterMC (talk) 10:40, 21 September 2015 (UTC)

Thanks for the great responses. As usual, the answers caused me to think of more questions.

Would making multiple hashes for the same data (with minor changes) make it easier to break the hash? For example, suppose I used the following inputs:

my email: foo@bar.com

My email: foo@bar.com

Contact foo@bar.com

foo@bar.com

Neglecting the fact that these are somewhat short, would using all of these at different times make it easier to extract the "foo@bar.com?.

How easy would it be to generate a different input to give the same output. This would be to enable "Plausible deniability" in the event that someone broke the hash, and tried to pin it on me when I don't want my identity known. Bunthorne (talk) 20:23, 21 September 2015 (UTC)

Yes. If you have a series of hashes, there is a higher chance that one of them could be reversed. But this is hard.

Finding a different input that reaches the same output is hard, but possible. md5sum is not completely safe against tailored attacks for people with sophisticated resources.

Still, just ask anyone to put name+email+some random text into a file and run "md5sum file." Let them generate different hashes with different random texts. That's pretty safe.--3dcaddy (talk) 00:10, 22 September 2015 (UTC)

Amazon ID

Does amazon (I use amazon.co.uk) have a unique id for each product? If so, how does one find out what it is and how does one enter it into the system to go to that item? -- SGBailey (talk) 23:35, 20 September 2015 (UTC)

Amazon Standard Identification Number. Vespine (talk) 01:57, 21 September 2015 (UTC)

If you go to the product page and scroll down to the box marked "additional information" you will see the ASIN. You can enter any ASIN in the Amazon search box to find any item.--Shantavira|^{feed me} 07:52, 21 September 2015 (UTC)

Thanks. That is sort of what I expected. It appears that for books the ISBN number is the ASIN. Amazons help says "You can search for an ISBN or an ASIN in our catalogue. If you know the ASIN or ISBN of the item you are looking for, simply type it into the search box (which can be found near the top of most pages), hit the Go button and, if the item is listed in our catalogue, it'll appear in your search results". So to test I copied B00P6JZ4UQ from a (uk) pair of shoes, went to the homepage, pasted it into the search bar and pressed the magnifying glass (presumably "search") (there is no "Go" button). And behold, I get "Your search "B00P6JZ4UQ" did not match any products". Is it them or me? -- SGBailey (talk) 21:52, 21 September 2015 (UTC)

Amazon doesn't find B00P6JZ4UQ for me either, but you can just use google, search for amazon B00P6JZ4UQ, that worked for me: http://www.amazon.co.uk/Clarks-Smart-Huckley-Spring-Leather/dp/B00P6JZ4UQ Vespine (talk) 22:49, 21 September 2015 (UTC)

September 21

Screen Reader shares…

Hello,

Some features in ʜᴛᴍʟ like ᴀʀɪᴀ are not directed to be supported by web‑browsers but by screen readers for blind people.

When I search if a feature is supported by screen readers, I see various lists.
However this doesn’t help because I have no idea about the traffic coverage, even for the English language.

Would it be possible to find such shares ? 2A02:8420:508D:CC00:56E6:FCFF:FEDB:2BBA (talk) 00:05, 21 September 2015 (UTC)

When it comes to Internet applications for the blind, you will likely not find anyone who knows more than Bryan Smart. I'd ask him directly. 209.149.113.66 (talk) 11:43, 21 September 2015 (UTC)

I know the stats for my language. But I need them for the English language in order to prove to a site a situation I discovered is harmful. 2A02:8420:508D:CC00:56E6:FCFF:FEDB:2BBA (talk) 18:19, 21 September 2015 (UTC)

Are you asking what percentage of viewers of a certain page are likely to be using screen readers? If so, this page has some links and discussion [11]. You can also just look at distribution of blind people. In the USA [12], I think it's safe to assume that most blind adults use the web, but that might be less true in other places. There also might be some useful info here at WebAIM [13] SemanticMantis (talk) 18:58, 21 September 2015 (UTC)

Not exactly, I found an English only site which don’t filter protocols on the cite="" attribute but is doing it for other ones like href="" data="" src="". They refuse to change the current behavior by claiming the percentage of blind users that would be affected by the javascript: protocol on the cite="" attribute is probably very small.

I found various screen readers vulnerable to that case, but I couldn’t found if there’re used by a lot of users. (and to be honest, there’s is a bounty to win if I prove such a situation is harmful for blind users)

That’s why I need screen reader shares, the same way browser shares exists. 2A02:8420:508D:CC00:56E6:FCFF:FEDB:2BBA (talk) 23:12, 21 September 2015 (UTC)

I'm not visually impaired, or that familiar with accessibility issues, but my impression is that JAWS (screen reader) is the most popular screen reader. This is supported by our article, with a source [14], although their percentage is now well below 50%. It seems from that source that JAWS, ZoomText, Window-Eyes, NonVisual Desktop Access, VoiceOver are what would need to be considered, at a minimum. (This doesn't tell you much about versions of these software of course.)

The source, a survey by WebAIM, doesn't seem to be exclusively referring to English usage statistics, but considering their are based in the US with an English website, and the survey was I think in English, and a big percentage of their respondents came from the US, follow by Europe (including UK), then Australia and Oceania, it seems likely English users were a big percentage of users surveyed. Although interestingly, 25.6% did use more than on language.

Being a survey, and from the sound of it not one that even attempts to be random, it does run the risks such surveys normally do, still it's probably better than nothing. Also coming up with a good alternative is probably difficult, since screen readers are generally not exposed in the user agent or elsewhere from the sound of it [15].

However are you sure the website in question doesn't have their own data? If it's a competently run major website, it may have. For various reasons, website audience will often vary, and while it's possible the website design itself is one of the causes, I think in the end many website designers will care much more about their actual audience than the general audience. (Browser shares, which are normally derived from user agents are I think often handled that way.)

If they do have their own data, and it suggests for example that VoiceOver is not particularly significant to them, considering the somewhat low usage share in the WebAIM survey, it unfortunately may be difficult to convince them it's something to worry about.

I'll also ask User:Graham87 who uses a screen reader and is often a helpful source of information for accessibility related issues such as this, if they can offer any more help.

Nil Einne (talk) 14:26, 22 September 2015 (UTC)

How could VW and Audi software cheat the US regulations?

According to [16] "Diesel cars from Volkswagen and Audi cheated on clean air rules by including software that made the cars' emissions look cleaner than they actually were."

How can the VW and Audi software detect the car is being tested? As a first thought, I imagined that they analyzed the emissions coming out of the exhaust pipe, and do not need to connect to the car board computer or look at the dashboard. At least, that's how I expect a technical inspection to work. Relying on the cars' own sensors or software is tricky, even without fraud. You would not discover software or sensors malfunctioning this way. --Scicurious (talk) 15:38, 21 September 2015 (UTC)

A comment [17] by 1995droptop claims "EPA testing rules do not allow for anything to be plugged into the OBD port during testing." Another comment by BahamaTodd says "The car was able to detect humidity, steering angle, vehicle speed, and duration of operation, to determine that it was undergoing an EPA emissions test." (Also mentioned is how the EPA generally uses fairly well controlled enviroments, I presume for consistency and fairness.) A search based on that finds other sources of varying quality like [18] [19] [20] which say something similar, and also go in to detail on what they did in the testing mode and why they may have chosen to change the settings in normal driving mode. One of those links to the EPA which calls it a "defeat device" [21] which doesn't go in to details on precisely how it's believed to have detected testing, nor does the Notice of Violation [22]. However, although not the best sources, I'm inclined to trust that the earlier ones probably right in how it worked. Nil Einne (talk) 18:18, 21 September 2015 (UTC)

(Full disclosure: I designed some of the equipment used to conduct the EPA tests in question.) Emission tests are normally done on a dynomometer using the FTP-75 test schedule.

All VW had to do was to program the car to go into EPA test mode when the computer saw that the vehicle was going through the FTP-75 driving cycle combined with none of the normal steering corrections that happen as a human stays in his lane on the highway.

It's a well-known problem in the industry, and we have detection measures which I am not at liberty to reveal. The interesting question is how VW managed to not trip those detection measures (which no doubt have been updated several times since I was working with them). --Guy Macon (talk) 04:37, 22 September 2015 (UTC)

How are Arabic glyphs actually rendered ?

Let's consider, for example, two letters: Arabic Jeem and the Pashto letter Dze. In Unicode the first has one its principal codepoint 062C, as well as four additional codepoints for its shapes (isolated, final, initial, medial) FE9D, FE9E, FE9F, FEA0. When you type the word jīm itself (جيم) the first glyph is supposedly not 062C but FE9F (the initial shape) this is why it has such a shape (ﺟ) and not (ﺝ).

The Pashto letter Dze (or Dzeem) in turn has only one Unicode codepoint 0681 and no codepoints for the shapes. But nevertheless the word dzīm is ځيم not ځ يم (I added a space between the letters to illustrate). How does it happen?

Bellow is the table. The first row is Jeem. It has four independent glyphs for its shapes. The second row is Dzeem. Its shapes are got with tatwil.

Final	Medial	Initial	Isolated	Character
ﺞ	ﺠ	ﺟ	ﺝ	ج
ـځ	ـځـ	ځـ	ځ	ځ

--Lüboslóv Yęzýkin (talk) 18:50, 21 September 2015 (UTC)

You’ll be interested in computer font kerning. It also allows to adjusts letters so they override. 2A02:8420:508D:CC00:56E6:FCFF:FEDB:2BBA (talk) 23:31, 21 September 2015 (UTC)‮

In Serbian, the italic shapes of certain letters differ from those in Russian. This is not encoded, the info has to come from somewhere else (such as the document's metadata - for example in wiki markup, the {{lang}} tag.) Similarly, in Greek, an algorithm to automatically put text into lowercase knows which form of the sigma to substitute (σ or ς) even though all it has is Σ (in other words, the relationship is one-to-many (non-bijective) and the algorithm has to guess based on context (such as whether the next character is a letter or punctuation.) The point is, Unicode text rendering engines are that smart Asmrulz (talk) 08:48, 22 September 2015 (UTC)

Spam e-mails

I receive quite a number of e-mails, ostensibly from people I know, usually sent to a number of addressees, with the heading "from: " followed by the person's name. The content is always just a web address (which I never click on) and the person's name. Should I be concerned about this, and should I notifying the supposed sender? Sorry if this is a well-known phenomenon that I ought to have known about! --rossb (talk) 08:17, 22 September 2015 (UTC)

There are many scenarios for this:

• Worst case... You have a virus/malware on your computer that is scanning your emails and sending you spam from people with whom you currently exchange emails.
• likely... Someone you know has a virus/malware on his/her computer that is sending out spam using all the email address in his/her email program.
• Most likely... A spambot is sending you tons of email using an email list harvested from some site you've given your email address to. Others who know also gave their email addresses to it. So, you sometimes see names you know.

Realization... The number of computers that have malware on them that spend all day and night sending out spam is MUCH larger that most people think. Simply getting everyone to turn off their computer when they aren't using it will prove this point because it will greatly reduce the processing power the spammers have. However, humans are highly addicted to screens and they don't like to have screens turned off. As a species, humans have become dependent on having screens scattered throughout the home, installed in the dash of cars, hanging randomly throughout stores and restaurants, on nearly every desk at work, and now they must carry a live screen with them everywhere they go and stare blankly at it while driving - only taking attention away from the mobile screen periodically to look at the other little screen on the dash... Seriously, I wasn't expecting this to turn into a completely off-topic rant. 209.149.113.66 (talk) 13:10, 22 September 2015 (UTC)