Jump to content

Wikipedia:Reference desk/Archives/Computing/2014 October 31

From Wikipedia, the free encyclopedia
Computing desk
< October 30 << Sep | October | Nov >> November 1 >
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.


October 31

[edit]

which is best Kurdish search engine?

[edit]

which is best Kurdish search engine? — Preceding unsigned comment added by Jarry jon (talkcontribs) 13:25, 31 October 2014 (UTC)[reply]

The Open Directory Project has a category in the Kurdish language (DMOZ - World: Kurdî), although it is not a search engine.
Wavelength (talk) 15:58, 31 October 2014 (UTC)[reply]
Google Language Codes - tomihasa has a link to https://www.google.com/webhp?hl=ku&gws_rd=ssl. You can click on the keyboard icon to enter search text in Kurdish characters.
Wavelength (talk) 16:11, 31 October 2014 (UTC)[reply]

An M-language?

[edit]

This is a typical Halloween story :-). I've been looking for FORTRAN code to calculate Associated Legendre Polynomials and Spherical Harmonics. There are thousands FORTRAN routines on the web. At one of the sites I found a zip file 888.zip which I opened in my Ubuntu VM. As a result I got a set of directories that have nothing to do with FORTRAN. Most files have .m extension and the source code appears to be written in some obscure M-language, if I am not mistaken.

I am wondering if anybody knows anything about this language and how to convert it to FORTRAN. My GFortran compiler could not handle the source code files. I hope StuRat will be able to crack the mystery :-)

Thanks, --AboutFace 22 (talk) 17:16, 31 October 2014 (UTC)[reply]

The zip file contains a data directory that states it is clearly a MATLAB program. See this for a blog about what is involved in converting MATLAB to FORTRAN. 209.149.115.7 (talk) 17:52, 31 October 2014 (UTC)[reply]

Thanks. It never crossed my mind. I've always thought MatLab online files are always in object code. It is a different tack. Thanks again. --AboutFace 22 (talk) 20:40, 31 October 2014 (UTC)[reply]

bash autocomplete disable?

[edit]

Every once in a while, perhaps due so something random I accidentally typed, filename autocompletion in bash stops working for me. I hit the tab key, and it just beeps and autocompletes nothing, no matter how much or how little of the fragmentary filename I've typed, no matter if there are 0, 1, or multiple matches. The only way I've found to fix this is to exit and reopen my shell, but this is a nuisance and loses context. Anybody know of a specific mode bash can get in that would do this, and how I can disable or otherwise correct that mode more cleanly? —Steve Summit (talk) 21:27, 31 October 2014 (UTC)[reply]

Which operating system and which bash version? This actually does make a difference: if you're using most Linux, you're probably using the real GNU readline as opposed to BSD's libedit-based readline. You might also be using the bash built-in readline code, as opposed to tab completion provided externally via the readline library. Knowing the version will help narrow the range of possible problems.
OS X contains some kind of GNU GPL readline as part of bash; it provides BSD-licensed libedit for everything else.
In normal use (depending on what you consider "normal"), you might actually be swapping between these two code-paths... that would throw a wrench into rl_complete (tab-completion). Nimur (talk) 22:04, 31 October 2014 (UTC)[reply]
It happened to me most recently (i.e. tonight) with "GNU bash, version 4.1.5(1)-release (i486-pc-linux-gnu)", on a Debian "squeeze" system. But I think I've seen it elsewhere, perhaps also on my Mac. (And, as for tonight's bug, I'm kicking myself, because despite intending not to I killed and reopened it, meaning that the failure mode is gone, and I won't be able to immediately test any suggestions y'all may have for me.) —Steve Summit (talk) 22:26, 31 October 2014 (UTC)[reply]
some distros (Fedora) hook (hack?) the completion feature to make bash autocomplete all sorts of things besides filenames, such as options to common programs (df --hu [tab] man-readable) and package names in rpm and yum (Fedora's package manager.) sometimes there's a noticeable delay when it tries to download stuff (list of packages) to know what to complete. also (I can't think of an example off the top of my head) I had cases when the autocompletion feature (wrongly) expected a long option instead of a filename in that position in the command line, and, not finding an option with the name as the file's, simply didn't complete anything. in Fedora, the bash-completion package is responsible for this Asmrulz (talk) 01:55, 1 November 2014 (UTC)[reply]
Good point about the hooks. In my case the commands I was using (that is, that the filenames I was trying to autocomplete were arguments to) were perfectly ordinary ones, like cd and ls. But perhaps a hook for some other kind of argument type had gotten activated previously and then, accidentally, not deactivated. —Steve Summit (talk) 16:17, 1 November 2014 (UTC)[reply]

Tor at college

[edit]

Can my college monitor what I do on tor? I know they can probably see the amount of data that is being transferred same as a normal ISP would see but I was curious if my college would be able to see anything I am doing. I feel like I am being spied on by the college and I want that to change. By being spied on I meant like a personal spying not just a normal NSA kind of spying. Also I live on campus at the dorm, they did not ever have access to my desktop to install software like I hear they do at some colleges.

Also, can someone explain to me how the "darkweb" works? I have heard things about being able to access like the full domain of a website and being able to see ever page but I do not really know how to explore the dark web. Are there any tools that I could use to map out a domain? Sort of like a visual map or in text telling me all the URL's. Thank you very much for your time. — Preceding unsigned comment added by 204.42.31.250 (talk) 23:38, 31 October 2014 (UTC)[reply]

Dark web. Deep web. Use with caution. I hope the college doesn't know that IP address.
You don't need physical access to install software.
Personal spying is the normal NSA kind. What other kind exists? (Nevermind. ELINT.)
If there was a tool for lighting up the world wide dark web, it'd just be the world wide web. InedibleHulk (talk) 00:45, 1 November 2014 (UTC)[reply]
If I'm correct, TOR encrypts all the data you send, so unless your college cracks the encryption, it's impossible to monitor the traffic you send/ — Preceding unsigned comment added by Kiwuser (talkcontribs) 01:59, 2 November 2014 (UTC)[reply]
As long as you're not using a school computer, didn't install school software, and as long as your school network permits use of proxies/Tor (which, although your specific activities wouldn't be visible, that you're hiding your traffic is possible to determine -- this is similar to how you may see that you're blocked from editing on Wikipedia if you're connected to a known proxy).
As people have already said above, use caution. Tor puts your information in a little package that stays closed as it's handed from one computer to another, with none of them knowing where it came from or where it's going...until the last node, called the "exit node" where your data is unencrypted and transmitted to the website you sent it to. But here's the thing: While the school may not be able to see what you're doing, you have no idea who is operating that exit node. You should assume that any information you transmit is visible, even if your location/IP is not. So I would never make purchases using credit/bank info, do anything requiring information like a social security number, etc. while connected to Tor.
Another option -- and usually faster -- is to use a VPN instead of Tor. Basically a internet proxy, but instead of an http proxy that only encrypts some web data, it creates a secure connection for all of your Internet use. There's still an unknown because you never know who at that VPN might be shady, but you have more information that way than you do with Tor. The reason people who are really into security prefer Tor is because with a VPN the operators of the VPN can connect your IP/identity/location to your activity, and they have varying policies about logging, whereas with Tor although what information you transmit may be visible to a sketchy exit node operator, your originating IP/location stays hidden. This doesn't really matter with the school unless they get a subpoena to have the VPN cough up logs. --— Rhododendrites talk \\ 19:34, 2 November 2014 (UTC)[reply]
Actually, there's no reason to assume that all information submitted over Tor is visible. If you're using end-to-end encryption (HTTPS in the case of the web) and you make sure the certificates are correct and your browser and the server is handling encryption properly etc, there's no reason to assume it is. If you aren't using end-to-end encryption, you probably shouldn't trust your own ISP, let alone uni access or a VPN, let alone a public access point (even if you're sure it's run by who you think it is and not an unknown third party); with credit card details etc. Remember that while it's unlikely in most countries anyone at the ISP or whatever is going to be allowed to official capture such details, you shouldn't assume the ISP has perfect internal security (which remember also includes routing over nodes they may have limited or no control over). Of course, history has shown it's not always wise to assume the company you're transmitting the info to has proper internal security anyway. That said, it's probably wise to assume unencrypted data over Tor is monitored, and also wise to assume there may be attempts to trick you in to using an unencrypted or improperly encrypted connection. So you should take extra care in what you do. Whereas it's perhaps not unresonable to worry about all this far less when simply using your ISP although as mentioned I'd still at a minimum, make sure the connect to the website is using HTTPS. Unfortunately with the move to apps on mobile device, unless you're monitoring traffic at the network level, it's usually difficult to know if the app is using a secure encrypted connection, even though many do simply use HTTP/HTTPS. Nil Einne (talk) 22:27, 2 November 2014 (UTC)[reply]


You might be interested in freenet, which is designed to be decentralized and fully encrypted. But yes, any ISP can monitor at least some of your internet activity, because they are in control of the pipes from your house to your first connection. So, while they probably can't tell what you're doing if you use a good VPN, tor, freenet, etc., they will most likely to be able to tell that you are using such products. This may show up on automated scans, and will certainly come to light if you are put under scrutiny. The usual defense for this kind of evasion is that e.g. journalists in repressive regimes should be able to report without censorship. If you are at college in the USA or EU, you most likely are not fighting a humanitarian crusade. If you just want pirate media, I urge you to consider why you deserve such things for free, and why you are willing to break laws to get them. SemanticMantis (talk) 15:44, 3 November 2014 (UTC)[reply]