Wikipedia:Reference desk/Archives/Computing/2011 June 14

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.

June 14

Browser war

I thought about it for a long time but could not find the reason behind the scene. All companies want users to use their own browsers instead of other's. But for what? They spend years writing a browser, then give it for free, not including many efforts to attract users. I do not believe they're kind enough to spend a huge budget for a software and give it away. The only reason I can think of is the default search engine. The default for IE is Bing, and Google for Chrome. Is there any other motivations after all? -- Livy the pixie (talk) 02:05, 14 June 2011 (UTC)

Exactly right, the advertising in search results is where they make money so market share is very important. (Correct me if I'm wrong but I think Google only makes money from advertising in it's products?) Another purpose is that it promotes brand loyalty from users. Benjamint 03:09, 14 June 2011 (UTC)

Not only advertising, but a fair amount is from that. ¦ Reisio (talk) 05:22, 14 June 2011 (UTC)

Years ago, Netscape had a paid version. Dismas|^(talk) 03:21, 14 June 2011 (UTC)

But Safari's and Firefox's default search engine is Google, too. Google pays Mozilla and Apple to develop Firefox and Safari, then? -- Livy (talk) 05:00, 14 June 2011 (UTC)

I think another strategy is to use the free browser to pull in customers for other products (for which they charge). StuRat (talk) 05:04, 14 June 2011 (UTC)

Mozilla gets a substanial amount of money from Google [1] Nil Einne (talk) 05:18, 14 June 2011 (UTC)

Web browsers are arguably the most popular applications used today, whether ordinary people can define what they are or not. Having the slightest bit of influence on people is incredibly useful. Just the default home page and bookmarks alone. The default search engine. Namebrand recognition for other products & services (consulting $$$). ¦ Reisio (talk) 05:22, 14 June 2011 (UTC)

Browsers are becoming a more and more important software platform, just as everyone knew would happen. The thought is that if any one browser got near 100% of market share, they could move away from the current open standards and develop their own value-add services that would only work on their platforms. (See Embrace and extend) This could conceivably be a good thing (open standards tend to progress slower than they could.) but it could also wind up being a sort of "Microsoft tax" where to build a web-site you are forced to buy some sort of Microsoft WebSite Kit. This sort of thing is common in closed systems controlled by a single entity. If you want to make an X-Box game you have to buy a special license from Microsoft, if you want to make an iPhone app you need to buy a mac.

In the extreme case, Giving control of web browsers to a single organization, could allow them to maneuver themselves into a position where they could phase out the web or certain aspects of it in favor of a network that benefits them. (Right now you can download many software products over the net. Imagine if the One Browser Manufacturer disabled this ability(citing security concerns) in favor of an iTunes-like app-store. Everyone else who sells software would be at a sudden disadvantage.) APL (talk) 06:37, 14 June 2011 (UTC)

Browsers do seem to be a case where competition has provided incredible benefits to customers. We now have a range of browsers, all of which are in pretty good shape, and they are all free. Nobody would even think of trying to charge for a browser now, due to all the free options out there. I only hope that some day this same description will also apply to operating systems. StuRat (talk) 07:08, 14 June 2011 (UTC)

Best Linux OS for File Transfer Protocol Server

Hello Everyone,

  I'm in need of a linux operating-system for an FTP server. The key requirements for the OS are:

1. The OS must have a graphical user interface.
2. The OS must be configurable to allow anonymous FTP access.
3. The OS must, once deployed, not need to be rebooted for package updates, upgrades, etc.
4. The OS must be stable, and not easily crash or hang.

  What are the best linux server OS which meet these criteria?

  Thanks as always. Rocketshiporion♫ 13:49, 14 June 2011 (UTC)

Any of them meet your criteria. Regardless of which you choose, you can opt for Gnome or KDE for your graphical interface (among many other less popular graphical interfaces), you can install FTP with anonymous access easily, you can ignore any future updates, and you have basic stability because you aren't using anything cutting-edge. All you really need to do is pick which package manager you prefer and use that to choose your flavor of Linux. I like yum, but some people hate yum. -- kainaw™ 13:58, 14 June 2011 (UTC)

The FTP server is just an application running under the operating system. Do you want a GUI to configure the FTP server itself? CS Miller (talk) 14:04, 14 June 2011 (UTC)

Yes, I want to configure the FTP server via a GUI. As for the package manager, I'm not at all familiar with any of the package managers used by Linux distributions. Rocketshiporion♫ 20:25, 14 June 2011 (UTC)

The general most common opinion, though far from universal, is that ubuntu is the easiest version of linux for an unitiated user to get started on, and install some of their favorite applications. Download and burn the install disc from ubuntu's website, put it in your drive, follow instructions. There's an application manager tool on the main menu. Use that to search for whatever FTP program looks best and install. i kan reed (talk) 21:08, 14 June 2011 (UTC)

I'll go with Ubuntu Server - it's downloading now. Rocketshiporion♫ 23:47, 14 June 2011 (UTC)

I've installed Ubuntu Server with the Samba file server option, but now the screen says rocketshiporion@ubuntu:~$ - where's the GUI, and how do I get to the main menu? Rocketshiporion♫ 01:12, 15 June 2011 (UTC)

Ubuntu Server is not Ubuntu. In general "server" means that a human will not be using the machine directly, so a GUI is a waste of resources. You want Ubuntu, not Ubuntu Server. -- kainaw™ 01:20, 15 June 2011 (UTC)

Ubuntu includes a whole bunch of packages and programs that are not needed for a simple FTP server (such as Firefox, Office, etc). If you go with Ubuntu Server it will not install any of those things, but also, it will not install a GUI. However, I believe, you can install a GUI on top of Ubuntu Server by logging in, and using the following commands (I haven't tested these myself):

sudo apt-get install x-window-system

followed by

sudo apt-get install gnome-core gdm

The other alternative is to install Debian. During the installation process, it will ask you what type of system you want. And you can choose to setup up an FTP server with a GUI. - Akamad (talk) 02:27, 15 June 2011 (UTC)

Also, a SAMBA server is probably not what you are looking for, as it is very different to an FTP server. I believe the FTP server on Debian (and probably Ubuntu) is called proftpd. - Akamad (talk) 02:38, 15 June 2011 (UTC)

If it isn't proftpd, it will surely be vsftpd. -- kainaw™ 12:08, 15 June 2011 (UTC)

backup hardlinks

Is there any file based backup software, except NTBackup, which supports hardlinks? Filesystem based backup would take care of hardlinks, but would be inefficient, because that volume contains some amount changing junk, which does not needs to be backed up (it would be backed up, because filesystem based backup programs operates on entire filesystems and usually cannot exclude files/directories). File based backup software, which does not supports hardlinks, either will deduplicate hardlinked files (backup will be normal size), or not (each hardinked file will take up number of times more space), but restore process, in either case, will create duplicate files, instead of hardlinks. In a filesystem with large number of hardlinked files, it could be impossible. This particular filesystem has size of 698GB, used space 605GB, and total size of all files (in windows explorer) 1120GB. -Yyy (talk) 14:20, 14 June 2011 (UTC)

Giant Mail Merge Question

Hello. I have a spreadsheet with over 400 records on it. What I want to do is set up a merge on Word so it will merge the data from each of these records into over 400 letters, each containing different information (according to the spreadsheet). This is all fine. Then, I will have a 400 page Word document. What I really want to know is if it is possible to then split this document into 400 separate 1 page PDF files. What is the easiest way to do this?

Thank you.

Jeremy Wordsworth (talk) 14:47, 14 June 2011 (UTC)

You can use PDFtk to "burst" a PDF into one-PDF per page: pdftk foo.pdf burst -- Finlay McWalter ☻ Talk 15:39, 14 June 2011 (UTC)

Learning game programming by reading open source code?

Hi, I'm practising game programming in C++ and I sometimes read code from open source games, such as Tux Racer or SuperTuxKart, to get an idea of how things are done in game programming. The games whose code I read are published under GNU GPL. However, I don't necessarily want to publish my games as open-source. Do I have to license them under GPL, even though I'm not copying the actual code but just the 'idea' behind it? 88.148.251.200 (talk) 15:06, 14 June 2011 (UTC)

No. --Sean 17:24, 14 June 2011 (UTC)

For a legal example and analysis of using ideas and standard practices in computer programming, see coverage of SCO vs. Linux coverage throughout the Internet and even here. -- kainaw™ 17:46, 14 June 2011 (UTC)

No. Code is copyrightable but ideas are not copyrightable. I salute you for doing this yourself. You must know this already but in case you don't, I'll mention that actually writing code (and debugging your own code) will teach you much better than reading code someone else wrote to do the same thing in roughly the same way. Comet Tuttle (talk) 21:10, 14 June 2011 (UTC)

Comet: you're absolutely correct but sometimes I've found clever ideas by seeing how other people have done things. It's good to get a basic foundation. HominidMachinae (talk) 01:28, 15 June 2011 (UTC)

Ideas are not copyrightable but they are, in principle, patentable. It is extremely rare for open source software writers to patent their methods, though. Looie496 (talk) 02:42, 15 June 2011 (UTC)

• Warning legal advice is not allowed on wikipedia I know since wikipedia is open source, we all consider ourselves experts on open source licensing but you may not give out advice on the reference desk regarding copyright law. Please seek advice regarding the legality of this elsewhere. i kan reed (talk) 14:24, 15 June 2011 (UTC)

Also as a follow up: some of what is said above conflicts with what I believe to be true under United States copyright law. Please do not take the above as legal advice. i kan reed (talk) 14:59, 15 June 2011 (UTC)

Well, I don't know any other place where I could ask this. 88.148.251.200 (talk) 16:54, 15 June 2011 (UTC)

You've never heard of http://www.gnu.org - which has an entire section dedicated to licenses? -- kainaw™ 17:01, 15 June 2011 (UTC)

As far as I have understood, in principle the GNU GPL says is that if you use GNU GPL licenced code in your own project and publish it, you must provide the source code for the GNU GPL licenced code. If you have not made any modifications to the GNU GPL licenced code, it's enough to say that you use the code and provide a link to the original source code. If you have made modifications to it, you must publish the modified source code. You never need to publish any source code to the parts of your project that are not GNU GPL licenced, even if your project uses (non-modified or modified) GNU GPL licenced code, only to the parts that are actually GNU GPL licenced. IANAL, but this is what I understand about it anyway. JIP | Talk 20:23, 15 June 2011 (UTC)

The only real caveate to using the "ideas" behind code is that some ideas are patented. See software patents. Some of what is patented matters quite a lot. With regards to GNU GPL, though, it is unlikely that any of the code would be patented. In general, though, the "ideas" behind code are free to be recreated; only the actual code is itself copyrighted and thus covered by the license. --Mr.98 (talk) 21:47, 15 June 2011 (UTC)

I've been reluctant to write this but I'm concerned that it hasn't been adequately explained. While it's true ideas can't be copyright movie plagiarism copyrighted (see Idea-expression divide) I think what 'i kan reed' is getting at is the distinction between an idea and a copyrighted expression of an idea (or a derivative work of that expression) is not always so simple. I don't know of many cases in the software world other then the infamous look and feel lawsuits particularly the Apple one which are more about visuals and interactivity then code. However there are examples in the literary world e.g. The Da Vinci Code#Lawsuits (and to some extent Legal disputes over the Harry Potter series#Adrian Jacobs and Tanya Grotter). Also in the movie world (e.g. the abandoned lawsuits I Now Pronounce You Chuck and Larry#Controversy) or the persistent controversy over the Lion King).

While most of these didn't succeed they give you an idea of some of the issues and also you will note in many cases the people denied there was any influence or they'd even seen the copyrighted work. I don't want to provide legal advice, but if it turns out large sections of your code is very similar to someone else's code, I wouldn't dismiss copyright violation as a possibility, particularly when you admit you looked at and were inspire by said code. Even though you may not be actively or intentionally copying the code or may not have read it for a while, how much are you going to remember and how closely are you likely to unintenionally follow the original code when programming? How many other ways were there for you to do what you did?

To use an example, I'm pretty sure Microsoft doesn't want any of their Windows programmers to actively work on or review the Linux source code (at least not officially) and the same for programmers for open source programs looking at the source code of highly related proprietary programs e.g. [2]. One of the reasons of course being even if they don't do it intentionally they may end up writing something closely enough to be at risk of a copyright violation.

Nil Einne (talk) 11:12, 16 June 2011 (UTC)

I'm the first person to agree that the idea/expression divide is a fuzzy and fishy one. However in the case of software, the practices we are talking about are quite common, and the use of patents in software is quite routine, unlike in the literary world. So I think it is a bit less fuzzy and less fishy for that reason — the idea/expression divide is much more codified in an engineering/programming context than it is a literary/creative context. But this is a subjective reading, of course, and not informed by legal training. --Mr.98 (talk) 14:49, 17 June 2011 (UTC)

worldwide visibility of jt3.com

At Talk:Area 51#JT3 Support we're discussing the accessibility of the website jt3.com. user:IHTFP can see it fine, as can downforeveryoneorjustme. But here in the UK I get no reply. For me, a traceroute goes no further than a machine registered to lvcm.com, a Cox Communications node apparently in Las Vegas (so I think the traffic is getting all the way to JT3's firewall). Given that the site's owner, JT3 LLP, is a US DoD contractor, I guess it may be blocking traffic from outwith the US. Can I ask other refdeskers, inside and outside the US, to check whether they can browse http://jt3.com/jt3.asp okay. Thanks. -- Finlay McWalter ☻ Talk 15:22, 14 June 2011 (UTC)

host-tracker is showing several countries where the site isn't available in 82.43.90.27 (talk) 16:49, 14 June 2011 (UTC)

Ah, thanks. That's just the kind of site I should have been asking about in general. -- Finlay McWalter ☻ Talk 16:58, 14 June 2011 (UTC)

Why can the site be reached from Norway and Poland, but not from Dallas, Texas? (or are these just glitches in host-tracker data?) Dbfirs 20:21, 14 June 2011 (UTC)

It's possible a glitche, maybe. I made up a completely random url and tried it, and 1 node out of 63 reported it was ok. 82.43.90.27 (talk) 21:45, 14 June 2011 (UTC)

Well we have no real idea why it isn't reachable from some places but not others. The idea they're blocking people outside the US is possible but it's really just a guess. Perhaps they've been a target for DOS or DDOS in which case they may block whoever has been D/DOSing them perhaps with some collateral damage but not those who haven't. Perhaps it's not intentional blocking but simply misconfiguration with one of the servers, in that case predicting who is going to be unable to reach the site will depend on precisely what the error is. For example, and this may also happen if one of the upstream routers is down, it may depend on what path you take to reach the main server which will depend on the routing configuration. If they really are intentionally blocking people outside the US, perhaps their geoIP database is very crap. Perhaps they have customers or potential customers outside the US in Norway, Poland and HK but none in the countries blocked. Or perhaps the blocking affects more than their web server, their email server or VPN or whatever and they have staff in these countries. You could probably get a better idea by doing a traceroute from various locations, finding other servers which pass through the first server/router that is unreachable (I don't know what this is since I can't access the site myself) etc but it seems a bit pointless to me. Best just email jt3.com and see if they fix it or just wait a few more weeks. Nil Einne (talk) 09:50, 15 June 2011 (UTC)