Wikipedia:Reference desk/Archives/Computing/2012 February 8

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

February 8

Joint authoring a document

Hello everyone. I live in NZ and am trying to co-author a paper with someone in the UK. We would like to both work on the same 'master' document (no forking or getting confused about which version we are working on; we especially don't want to email versions back and forth). We find 'track changes' or similar to see what the other person has added or changed useful. I use a macosx and kubuntu with openoffice, he uses windows with office. I am not sure he could cope with svn. Can any refdeskers advise? Thanks, Robinh (talk) 03:02, 8 February 2012 (UTC)

Google Documents supports collaborative editing, and has some support for revisions (I haven't worked with revisions much, so I can't tell how well-implemented that's it). Paul (Stansifer) 03:06, 8 February 2012 (UTC)

thanks! Looks good to me. I didn't know googledocs had a revision history.

Resolved

Robinh (talk) 03:31, 8 February 2012 (UTC)

computer project

I want to lern about computer project .Please tell me different project nane and there fild. I am Account student in Gujarat University and i want make my own project on Account fild .I also know about managment fild project. — Preceding unsigned comment added by Niravbhavsar (talk • contribs) 05:06, 8 February 2012 (UTC)

I have a hard time understanding you. If you are asking about computer programs in the accounting field, then spreadsheets come to mind immediately. Follow that link to read more. StuRat (talk) 06:29, 8 February 2012 (UTC)

Greater resolutions in Virtual Box

Resolved

I'm running Ubuntu 10.4 in Virtual Box. I'd like a resolution above 800x600 but I don't have any options above that. I'm running the VM on a Lenovo ThinkPad T400. How can I get a larger resolution? Thanks, Dismas|^(talk) 10:41, 8 February 2012 (UTC)

Have you tried assigning more memory in the display->video settings tab? Also, have you installed the VirtualBox ubuntu guest additions (I don't know about a Linux guest, but at least for a Windows guest that installs a more suitable video driver in the guest). -- Finlay McWalterჷTalk 11:00, 8 February 2012 (UTC)

In my experience (Ubuntu and CentOS on MAcOS-X) you always need the Guest Additions for proper video support, including fullscreen mode.--Stephan Schulz (talk) 12:32, 8 February 2012 (UTC)

And mine: sudo apt-get install build-essential module-assistant && sudo m-a prepare then mount guest additions (Install guest additions from VBox menu), and run the Linux installer shell script. ¦ Reisio (talk) 13:32, 8 February 2012 (UTC)

You also need to ensure you actually install it, rather than running live. It's really easy to run live off the CD and think you've done an install. --Phil Holmes (talk) 16:59, 8 February 2012 (UTC)

(OP here) The guest additions got it. Thanks everyone!! Dismas|^(talk) 20:11, 8 February 2012 (UTC)

Google Documents Versioning

As mentioned above, Google Docs includes a versioning facility, although it works slightly differently for Google's native formats and docs saved in their own forms.
I'm trying to save a new version of an Excel spreadsheet. I've done so several times before, but now, it insists on creating a new document with the same name. Can anyone suggest why it's suddenly behaving differently?
Thanks, Rojomoke (talk) 14:12, 8 February 2012 (UTC)

This suggests that true versioning is only for Google Docs format documents, and that for non-Docs documents one saves multiple versions, as it's asking you to do. -- Finlay McWalterჷTalk 15:57, 8 February 2012 (UTC)

Thanks, but that's what I meant by "works slightly differently for Google's native formats". For non-Docs documents, you save multiple versions, but they should appear in the file listing as a single item. I'm using "Add or Manage Versions" option to add a new version, but it's appearing as a completely separate entry. I hope that clarifies the problem. Rojomoke (talk) 16:51, 8 February 2012 (UTC)

Linux command: dir <number range>

I have a big list of files with number as names. How do I get a list of files number x to number y or higher than z? 188.76.164.174 (talk) 14:14, 8 February 2012 (UTC)

Are the names like 1, 27, 316, or are they like 0001, 0027, 0316 ? -- Finlay McWalterჷTalk 14:25, 8 February 2012 (UTC)

They are time stamps like 20111231134533 (meaning year, month, day, hour). Higher than x seems easy 2011* will deal with that. But I don't know how to choose a range. — Preceding unsigned comment added by 188.76.164.174 (talk) 14:28, 8 February 2012 (UTC)

I don't think ls will do that itself; you'd need to wrap it either with grep or with a bash script. Thinking in bash hurts my brain, so I can't help you with that (but it's surely a one-liner). I can easily give you a python script that will do it, if you like. -- Finlay McWalterჷTalk 15:02, 8 February 2012 (UTC)

You can include regex ranges in the arguments for ls, for example: ls -l 201112311345[30-33] Keegstr (talk) 18:13, 9 February 2012 (UTC)

That doesn't do what you imply: [30-33] (in glob or regexp) is the same as [0-3] (because it expands to [301233]). Of course, in that case you can simply write 3[0-3]. The easiest general solution is to use awk:

ls | awk '$1>=x && $1<=y {print}'

--Tardis (talk) 04:21, 10 February 2012 (UTC)

Server load estimate

For an auction website which has data that is regularly updated on various pages (e.g. item page, home page, users' watchlists), I would like to figure out a relation between number of users, number of pages being refreshed (or number of values?) and refresh interval such that the server does not break. Basically what I would like to know is, how often and on how many pages may I refresh for say 100 visitors to the website? Could somebody point me in the right direction? bamse (talk) 16:55, 8 February 2012 (UTC)

To do this analysis I would create a mockup of the auction page so I could figure out which data, and then how much data, will be refreshed; then you can do a little math to figure out the answer. Without the mockup, any guessing may be useless. To perform this analysis quicker you might want to base the mockup on an existing auction website. Comet Tuttle (talk) 18:49, 8 February 2012 (UTC)

If you're looking for a "mockup" of a commercial website, you don't have to start from scratch. The "canonical" example of an enterprise server software suite is the Java Enterprise Edition Pet Store, a tutorial and sample code that runs on Java EE. The sample program implements a full AJAX-enabled web front-end, a server and database system back-end, complete with load-balancing. Most of the code is free software; it runs on Apache Tomcat, (it looks like newer versions are even more tightly integrated with GlassFish), or you can run it on your own favorite Java application server. Nimur (talk) 19:48, 8 February 2012 (UTC)

Thank you for the replies. For the estimate does it matter whether:

1. N users are watching the same page (i.e. does it count N times or is it the same as if just one user watched that page)?
2. a page contains only one variable being refreshed or more than one (do the type of variables matter)?

bamse (talk) 16:26, 9 February 2012 (UTC)

Of course, all of these things matter - but any individual item's impact to total system performance depends entirely on your server software architecture. These systems are very complex, and usually require a lot of professional, highly-trained engineers who spend a lot of time to design and analyze them. You might start by reading Profiling an Enterprise Application in NetBeans IDE. Run through the entire tutorial. NetBeans is a very powerful, free, open-source software development environment suitable for tackling this problem. It can be used, for example, to generate graphs like this, showing CPU and Memory and other resource utilization. Nimur (talk) 19:08, 9 February 2012 (UTC)

Thank you. I will do some reading now. bamse (talk) 08:03, 10 February 2012 (UTC)

Cell-phones' OS

Which OSs do cell-phones use/d normally? (excluding the obvious, like Android, Symbian, RIM's and such). — Preceding unsigned comment added by 88.8.79.238 (talk) 22:33, 8 February 2012 (UTC)

I don't know, but your question seems curious to me. Won't the ones "normally" used usually be the "obvious" ones? 86.179.114.39 (talk) 22:43, 8 February 2012 (UTC)

I mean, not long ago, the OS was not explicitly stated by the manufacturer or re-seller. You just got a Nokia this or a Sony-Ericson that. Which OSs do or did these cell-phones had? 88.8.79.238 (talk) 23:17, 8 February 2012 (UTC)

In general, such devices are called embedded systems, and are often powered by commercially-available operating systems such as μCOS, eCos, ThreadX, rtxc, QNX, linux (in many forms), or a Java ME VM system. Perhaps they were powered by a similar, but not-generally-available-for-purchase operating system, in which case this portion of the software system may not have had a marketing name. These sorts of operating systems are very familiar to embedded system programmers; they have the sorts of tools and utilities that are helpful to a programmer who is designing a cell phone; but they often do not have the sorts of things an end-consumer expects from an operating system (like a user interface). For example, on uCOS, it is not even possible to "run a program." (There's no shell, and no fork; just threads, called "Tasks"). The operating system exists only to assist the system programmer in creating a complete product: it provides facilities for management of system resources, like memory, I/O, sharing the CPU, and so on. Nimur (talk) 00:07, 9 February 2012 (UTC)

(treading carefully so as not to ask a dumb question) Sooo, in the article on the Nokia 3310 why is there 'n/a' in the section 'Operating system' in the infobox? Is it just some primitive embedded software that runs the unit? --Ouro (blah blah) 07:10, 9 February 2012 (UTC)

A simple(r) embedded system will typically run a micro executive, which only does a handful of basic things (threads, queues, message-exchanges, semaphores, mutexes, ISRs). That's often nothing more than a library file. That's compiled with the monolithic system software (hardware bring-up, device drivers, and the actual phone application) all to produce a single statically-located binary image (often as a Motorola S-Record or Intel Hex format) ready to be loaded onto the system's PROM. In cases like that there's no isolation (the CPU/MMU doesn't support memory protection anyway) and the "API" that "applications" call is nothing more than a header file which makes direct calls (not syscalls). Applications (like the snake game) are just compiled in with the rest of the system firmware, and changing or adding one requires recompiling the whole thing. This obviously poses low requirements on the system, but it's obviously inflexible, so mobile phones have graduated to more sophisticated systems with proper isolation, real APIs, and separate linking, that we can reasonably compare to the OSes of desktop machines. -- Finlay McWalterჷTalk 12:01, 9 February 2012 (UTC)

From what I've read and seen, most of the cheap Chinese phones using MediaTek chipsets use Nucleus RTOS. Both articles also suggests likewise (although it may be a specific variant of Nucleus RTOS by MediaTek called MAUI). The Nucleus RTOS article suggests some non Chinese phones also used it. [1] while coming from someone involved with the company making Nucleus RTOS has some discussion in the comments on OS usage. (Note as mentioned in the comments and our Nucleus RTOS article this can be a bit confusing because the phone may use a RTOS for the baseband processor but a different OS for the application processor.) Nil Einne (talk) 14:56, 9 February 2012 (UTC)

And, of course, TRON. ¦ Reisio (talk) 14:09, 9 February 2012 (UTC)

The Nokia 3310 probably runs Series 30 (software platform). 67.117.145.9 (talk) 18:52, 9 February 2012 (UTC)

While I at first smiled when a direct answer like that was given, looking at the article for Series 30, it says the 3510 was an early Series 30 phone. Nokia 3310 predates that by three to four years, so I'm inclined to think it runs on an earlier platform. --Ouro (blah blah) 10:13, 11 February 2012 (UTC)