Wikipedia:Reference desk/Archives/Computing/2012 August 15

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.

August 15

I am Naveen Ganesan

how i insert an image in user page...? pls anyone help me in clear steps....? — Preceding unsigned comment added by Naveen Ganeasn (talk • contribs) 03:28, 15 August 2012 (UTC)

Please ask this question at the Help desk, WP:HELPDESK. That is the place for questions about how to use Wikipedia. Looie496 (talk) 04:09, 15 August 2012 (UTC)

Regex and sed

I have an html file that has a table in it. The table has a number of entries such as "S1234A56" or "S123AA45". They all start with S, have 3-4 digits, 1-2 letters, and a final 2 digits. They are all exactly 8 characters long. I need to change all of them so that the string is put into a URL. I was hoping to just use a quick Perl script on my Linux machine to make all the substitutions but I'm stuck on how to write the sed command. Will I have to use escapes for all the non-alphanumerics in the URL? Or can I wrap it in something that will pass it without escapes? Right now, I've got something along the lines of

sed "s%S\d\{3,4\}\a\{1,2\}\d\{2\}%<a href="https://www.someurl.com/foo.pl?id=MYSTRING>MYSTRING</a>%"

And replacing MYSTRING with the regex from the first part. The reason I'm using % for the delimiter is that it doesn't appear anywhere in my URL.

So, where am I going wrong? Is there an easier way that doesn't involve sed? Thanks for any help, Dismas|^(talk) 04:15, 15 August 2012 (UTC)

From my initial investigation sed does not recognise \d as meaning a digit. My version of sed thinks that is just a letter d. Therefore you will need [0-9] instead and you need the \( to select the string to replicate - so

$ sed 's%\([0-9]\{3,4\}[A-Za-z]\{1,2\}[0-9]\{2\}\)%<a href="https://www.someurl.com/foo?id=\1">\1</a>%g'
1234TT77
<a href="https://www.someurl.com/foo?id=1234TT77">1234TT77</a>

Thanks! That worked perfectly! There were about a half dozen that didn't fit the scheme because they were labeled things like LOANER1, LOANER2, etc. But they were easy enough to do by hand. I'd have hated to do all the rest by hand since there were a few hundred. Thanks again! Dismas|^(talk) 10:07, 15 August 2012 (UTC)

zero indexing of arrays

Hi, this is kind of a quirky one, but I'm doing some massive array storing (and manipulating, but mainly storing) using Strong's Concordance. All entries are stored with a number, and the numbers start at 1. As a result, I'm using an array where the first (ie, the "zeroeth") entry is blank, so the array index number matches the Strong's concordance number. Is this considered good coding practice? Is there some set of rules stating when you should do this, when you should avoid it, and when it's basically six of one, half a dozen of the other? IBE (talk) 07:24, 15 August 2012 (UTC)

There's nothing wrong with it. What you start your array with — heck, what keys you use in general, since that's entirely flexible with associative arrays — is just a matter of convention and keeping things internally consistent. Just make it clear that you're using one-based indexing and all is well. Depending on your language, it's probably absolute bestest practices not the initialize the zeroth array entry anyway (why use extra memory?), but really, it won't matter as long as you never try to do anything with it. --Mr.98 (talk) 13:01, 15 August 2012 (UTC)

I don't know why Mr.98 assumes you're using an associative array. I'm going to assume you're writing in C. Since you're using C, the advantage of having zero-based arrays is that (array identifier) + (index of element) = (address of element): see Array data type#Index_origin, and more at Zero-based_numbering#Advantages - another advantage is that you can use the modulus operator (%) to do certain array-navigating tasks conveniently.  Card Zero  (talk) 16:57, 15 August 2012 (UTC)

I'd say it has a minor whiff of code smell but it's reasonable in the situation you describe. Numerical Recipes in C follows a similar convention, so that the C programs use the same subscripts as the corresponding programs in the Fortran edition of the book (Fortran uses 1-based arrays). 67.122.211.84 (talk) 22:00, 15 August 2012 (UTC)

The code is in JavaScript, but for most purposes, simple array behaviour is much the same as in C, and the syntax is of course C-style. Yet js is such a messy language that it requires the programmer to watch his every step. I have function names almost as long as the screen, with information like fileNamePurposePrimaryInterfaceCreateListOfSuchAndSuch - all just to avoid confusing myself. Hence the concern about getting the array indexing as neat as possible - js is very good at allowing you to stuff up. I'll stick with the current system, but would be interested in further discussion. I don't have any need for accessing the address (I don't even know how to do it in js), and there is no modular arithmetic - it's just the big file of one dictionary, with entries numbered 1 to about 5400. Hence for this one big thing, it made sense - Strong's concordance numbers are a key part of Biblical scholarship, so I tend to think in terms of the number, and it would be hard to change. Still, I am open to persuasion - I've kept this kind of indexing out of every other part of the program, because I can see the danger. IBE (talk) 13:09, 16 August 2012 (UTC) Further point: I'm pretty sure it's an associative array - I don't know what a dissociative array would be. It's a standard array created by the Array() constructor in js. IBE (talk) 13:12, 16 August 2012 (UTC)

If you're using javascript, leaving gaps in the array is fine (apparently there's a gap of 100 entries in the middle of Strong's concordance anyway, so leaving entry #0 empty is insignificant beside that). I can't find a good reference for this, though. You can have a bad reference, if you like (i.e. what some guy on the internet says): [1] look at the reply to the second answer - "all JS implementations have 'efficient' implementations of arrays, at least they are substantially more efficient than normal objects. They all assume that a compact array is the norm, and a sparse array is uncommon, and for the reason optimise for 'normal' array usage." So your array should be compact, which is not the same as being completely free from gaps.  Card Zero  (talk) 14:48, 16 August 2012 (UTC)

Thanks to you all for the help, IBE (talk) 19:12, 16 August 2012 (UTC)

I usually prefer to start arrays at 1, even if that wastes the zeroth one. Reasons: the first one is 1, the second one is 2, the nth one is n. Bubba73 ^{You talkin' to me?} 20:22, 16 August 2012 (UTC)

Yeah, for javascript, don't worry about 1 or the gaps. 5400 is a small array by today's standards. Massive would be billions or at least millions. 66.127.54.117 (talk) 07:30, 18 August 2012 (UTC)

Variables in wikipedia

Are there any plans to support variables in wikipedia? Say, if I have a number which appears in several wikipedia articles and this number changes, ideally I would only edit it in one place, not in all articles where it appears. (I just updated Lists of National Treasures of Japan with three new designations and had to change numbers in several places.) Or would this feature be too complicated for the users? bamse (talk) 08:36, 15 August 2012 (UTC)

If you need help using Wikipedia, you'd be better off at the Help desk. Future plans for Wikipedia can be suggested and discussed at the Village pump. There appears to be a limited implementation of what you have in mind, known as Magic words - see also the MediaWiki 'Magic Words' help page. This seems too be only useful for things like current time and date, though. The text on the first page suggests that there's a server cost in calling the value of a variable, so it may be that having lots of variables that need to be queried for the current value each time a page loads (if that's how it works) would be prohibitively demanding on the servers. But you can make the suggestion at the village pump - I'm sure someone will be able to tell you exactly why the sky will fall in if it's adopted :-) - Cucumber Mike (talk) 09:48, 15 August 2012 (UTC)

You could also use templates, so for example a template called template:numberoftreasures and then on this page put your number. Then insert this variable with {{numberoftreasures}} However it does also make it more tricky to update and more likely a person will make a mess. Also if the page becomes too complex it may not render. Graeme Bartlett (talk) 10:57, 15 August 2012 (UTC)

Thanks for the replies. bamse (talk) 11:36, 15 August 2012 (UTC)

If you want an example of the "may not render", you can check out List of hard rock musicians (A–M) and List of hard rock musicians (N–Z). It was originally one list which I split into two. Part of the reason why it was me who did it was that when the article was brought to mine and other editors' attention, I was one of the few who could get it to render completely. The roughly 800 references were bogging it down. Dismas|^(talk) 18:30, 15 August 2012 (UTC)

I proposed a way to incorporate variables (birthdates, temperatures, stockprices) into MediaWiki without using templates, using a Facts tab instead, making it very easy for robots to maintain such data. Unfortunately the project doesn't seem very active. Joepnl (talk) 20:31, 15 August 2012 (UTC)

Address bar search (email id)

Like many other people I also search directly from address bar (Ctrl+L→Type→↵ Enter). Today I put an email id in address and presed enter and I got message "You are about to log in to the site "gmail.com" with the username "abcdef", but the website does not require authentication. This may be an attempt to trick you. Is "gmail.com" the site you want to visit?" – is this an expected behaviour? --Tito Dutta ✉ 09:12, 15 August 2012 (UTC)

Yes. If you use an URL with an "@" in it, it means, "try to log in as this user." It's a bit antiquated — it's not using Google's login, for example, but .htaccess authorization. This is usually only used for doing things like logging into servers, routers, or really quite old systems; it's rarely used for actual sites because the name and password are usually sent in the clear (that is, not encrypted in any way). The "tricking" aspect is because it used to be (in the mid-1990s) a common phishing tactic to obfuscate URLs this way. So you'd put in something like http://youractualbank.com?toomanyqueryvariablesforyoutonoticethatactuallywe'regoingto@myevilsite.com. It would thus be hard to see which site you're really visiting — that whole first part is just being passed as a "user name" to the actual site. So now browsers usually warn you if something like this is happening, because people rarely try to log in using this anyway. --Mr.98 (talk) 13:10, 15 August 2012 (UTC)

The above is basically what Safari gave me when I plugged in an email address. I got a warning saying that I was trying to access a potentially dangerous phishing site. Especially funny considering I used my work email address for the test.  :) Dismas|^(talk) 18:18, 15 August 2012 (UTC)

Except that the quoted URL actually works and lands you at 'youractualbank.com' instead of a phishing site :) Unilynx (talk) 19:57, 15 August 2012 (UTC)

Sorry, yeah. The description about phishing scams was closer to what I got. Dismas|^(talk) 20:23, 15 August 2012 (UTC)

Try it with quotes surrounding the address. ¦ Reisio (talk) 21:59, 15 August 2012 (UTC)

 Great! Quotes surrounding it a great idea! --Tito Dutta ✉ 13:18, 16 August 2012 (UTC)

Audio file on facebook

How to upload Winamp media file on facebook from computer? It is an audio file, not video file. Sunny Singh (DAV) (talk) 09:12, 15 August 2012 (UTC)

There are some apps like this, you can search in Facebook apps centre more. Also does it work when you try to attach it (in messages)? --Tito Dutta ✉ 09:15, 15 August 2012 (UTC)

PHP, Replacing Arrays

I have a php script that updates arrays after each execution. They are values that belong to a group, when the script is executed those variables may change to another group.. so I have group 1, group 2, group 3 & group 4. So when the php script is executed it gets more values (new values sometimes) and the group for each one. So if there are new values, it just appends them to the array, for the old values I need to remove them from the old group and move them for the new group. Take this example: There are 3 baskets with fruits already.

Extended content
$previous=Array ( 'Basket1' => Array ( 'berries', 'pineapple' ), 'Basket2' => Array ( 'red apples', 'Strawberries' 'melon' ), 'basket3' => Array ( 'avocado', 'peach', 'pear', 'green apples', 'banana' ) )

When the scripts run again it updates the info, with new fruits, or/and fruits in different groups, so I have create an array that contains the updated information+the old information...

Note:in this example there are 3 baskets in the real code there's actually a lot more arrays.

The thing is... I exactly don't know how to update the info... I could manually make a lots of 'ifs and for loops but it doesn't seem right, and is not extensible. I created this function to delete old values that appear on another groups (The fruit has changed bascket).

Extended content

function RemoveChangedValues($oldBasket,$secondOldBasket,$basket,$secondBasket){//whatever return Array( array_diff($oldBasket,$basket), array_diff($secondProxy,$forthProxy));//this returns an Array of the "old basket" and the "second basket" with the repeated values removed //later the both arrays (old and new) can be concatenated and then remove the duplicate values of each array... or.. that's the idea. }

The thing is that I can't find a way to easily loop through all the groups and call that function (other than writting lot's of if statements). Any ideas how this can be done? thanks! 190.158.212.204 (talk) 17:10, 15 August 2012 (UTC)

Can you expand on "When the scripts run again it updates the info, with new fruits, or/and fruits in different groups, so I have create an array that contains the updated information+the old information..." ? How exactly do you get the info telling you which fruits to add to each group, remove from each group, and move from one group to another ? An example would be good. StuRat (talk) 04:15, 16 August 2012 (UTC)

When the user submits some data, it sends information, it's converted into an array, I got the array from previous inputs, but I don't know how to merge the arrays. Is not actually 'removing' them it's just moving them, (removing then adding), you see if at the first time, I get 'apples' in the first group, then get the same item 'apples' but in the second group, then I would have to remove apples from the first group because it has moved to the second group. Is not anymore on the first group. The only thing the script gets it's fruits and their groups, I have to merge them with their previous results... — Preceding unsigned comment added by 190.158.212.204 (talk) 04:47, 16 August 2012 (UTC)

What I'm trying to get from you is the format the information is supplied to your script. Is it like this:

APPLES  -> BASKET1
PEACHES -> BASKET2
GRAPES  -> BASKET1

Or like this:

BASKET1 -> APPLES, GRAPES
BASKET2 -> PEACHES

Or something else ? StuRat (talk) 05:21, 16 August 2012 (UTC)

In any case, it would certainly help if you could refer to arrays by number, rather than names like "Basket3". I don't use PHP, so I'll use generic pseudocode. Does it support an array of lists ? If so, you could make that "Basket[3]", which would make loops work better. Something like:

FOR B = 1 TO HIGHEST_BASKET_NUM
  FOR N = 1 TO HIGHEST_FRUIT_NUM_FOR_BASKET[B]
    IF (BASKET[B].FRUIT[N] = CURRENT_FRUIT) THEN
      DELETE (BASKET[B].FRUIT[N])
    ENDIF
  ENDFOR
ENDFOR

If PHP doesn't support an array of lists, then you could just make it a two dimensional array:

FOR B = 1 TO HIGHEST_BASKET_NUM
  FOR N = 1 TO HIGHEST_FRUIT_NUM_FOR_BASKET[B]
    IF (FRUIT[B,N] = CURRENT_FRUIT) THEN
      DELETE (FRUIT[B,N])
    ENDIF
  ENDFOR
ENDFOR

(Note that my loops above start at 1, since I assumed the arrays start with the first element. If they start with a zeroth element, then the loops should start at 0.)

However, at some point this approach would become unwieldy, say if you have thousands of baskets each containing thousands of fruits, then this would take millions of checks to find any given fruit. In this case, you might want to create an alphabetized index, which would store each fruit, along with the basket it occupies. You could then do a binary search to find the fruit in the index, and that would give you the basket. (You could also index it's position in the basket, but then you'd need to update this whenever any lower fruit was deleted, assuming you repack the baskets to take up empty spaces.) The index would still need to be updated when a fruit is moved, which would be fast, or when a fruit is added, which would be slow.

To give more specific advice, we'd need to know the max number of baskets, max number of fruits in each basket, number of updates expected, etc. StuRat (talk) 05:46, 16 August 2012 (UTC)

Monitor problems

Hello! I'm having a strange problem with my monitor. Whenever I turn it on after not using it for a while (i.e. overnight or when I return from work), it does one of these: 1) Screen flickers rapidly on and off for a few seconds and then stays on like normal; 2) Flickers the screen on once, then immediately off again, and I have to push the on switch twice (on -> off -> on) for it to try again--this can happen up to four times before behaving normally; or 3) A combination of #1 and #2. Once the screen stops behaving strangely and is in the "on" state, it behaves normally and experiences no more problems for as long as I keep it on. Any ideas as to what is causing this? -- 143.85.199.242 (talk) 17:34, 15 August 2012 (UTC)

LCD? CRT? ¦ Reisio (talk) 21:56, 15 August 2012 (UTC)

-- Sorry; a Samsung LCD, SyncMaster 226BW. (I am the IP.) -- Tohler (talk) 04:00, 16 August 2012 (UTC)

Sounds like a classic cold start problem. It might be that the temperature is the problem, and it needs to warm up a bit for the thermal expansion to bring two components into constant contact with each other. Does this also happen if it goes into standby mode ? If not, then just leave it on standby. If standby mode doesn't fix it, then you're down to leaving it fully on all the time, putting up with this annoyance, or replacing it. I doubt if it would be financially viable to have it fixed, unless it's a very expensive monitor. StuRat (talk) 04:07, 16 August 2012 (UTC)

What are the consumer protection law (aka lemon-laws) like in your area? (I think you are in Arizona, US) If the monitor is still covered by them, I'd claim under those. The symptoms sound like a fluorescent lamp failing to start (ballast/starter on its way out). Do you know what kind of backlight the monitor uses? If you put an image with large black and white areas on it up, and then turn the monitor off, and the back on without it starting correctly, can you just make out the black and white areas? CS Miller (talk) 14:55, 16 August 2012 (UTC)

Csmiller: I created an HTML file with a black BG, containing tables with white BG interiors and black text within. All of it was completely visible as the monitor flickered.

StuRat: I never use standby mode; I always shut my monitor off when I'm not going to be using it for a long while.

I neglected to mention that I have been using this monitor since mid-2010 with no problems; this began within the past month. -- 143.85.199.242 (talk) 15:29, 17 August 2012 (UTC)

Well, try standby mode to see if it fixes the problem. StuRat (talk) 15:38, 17 August 2012 (UTC)

Help with Changing Youtube Email

I have a rather annoying problem and I hope you don't mind helping a non-tech savvy person like me out. I have a G-mail account that I use for private email with my personal name, and I have linked it with my Youtube account. I want to set up a new G-mail account with the same name as my Youtube account so that I can give it to people so I can talk with them without giving my personal name out. However, when I tried to set up the G-mail account, it won't let me because it says someone else has that name (Probably because my Youtube account has that same name.) What can I do to set up the E-mail account I want? Rabuve (talk) 20:13, 15 August 2012 (UTC)— Preceding unsigned comment added by Rabuve (talk • contribs) 18:36, 15 August 2012 (UTC)

I wouldn't get too optimistic, but take a look at http://support.google.com/youtube/bin/answer.py?hl=en&answer=183819 ¦ Reisio (talk) 21:54, 15 August 2012 (UTC)

And the second part, you can not sign up for that Gmail address too. I have some suggestions still– do you have lots of videos in the channel? 2) have you tried Google Apps for the channel's name (Google Apps is free, but you need to have the domain), is the domain name available? --Tito Dutta ✉ 13:22, 16 August 2012 (UTC)

I tried setting up a slightly diferent email account (my youtube name plus a few numbers) but now it won't let me add it as an alternate email. Any advice? Rabuve (talk) 18:39, 16 August 2012 (UTC)

How much would this storage server cost?

A recent article at MBNet.fi mentioned this storage server, saying it could hold a maximum of 32 terabytes and prices start at 400 €. 400 € seems awfully low for 32 terabytes - that's only a bit more than 1 cent per gigabyte - so I'm guessing it's for the server only, without any disks. How much would such a server fully laden with 32 terabytes of storage cost? JIP | Talk 21:12, 15 August 2012 (UTC)

It has 4 bays each for a single SATA3 disk; the list of compatible disks has at most an Hitachi 4TB disk. So that's 16TB. That disk, an Hitachi-HGST Ultrastar 7K4000, retails in the UK for about £450. I don't see a listing for a compatible 8TB disk, which would need to bring it up to 32 TB. I imagine Hitachi et al will have an 8TB in the works; you can extrapolate a bit and get maybe £1000 for that, when it's released. £4000 is about €5000. -- Finlay McWalterჷTalk 21:46, 15 August 2012 (UTC)

Disks are pretty obviously not included in that price. 4TB drives are still exotic (and reportedly not that reliable) and 8TB drives are probably still at least a couple years away. Note also you shouldn't set up a disk array for any important read-write usage without some type of RAID. With four 4TB drives, you'd get either 8TB (RAID-1 or RAID-10) or 12TB (RAID-5) of usable capacity. 4TB Hitachi Deskstars are around 270 USD (jr.com, newegg.com) last time I looked, and the faster and maybe more reliable Ultrastars are around 450 USD. 2TB drives are somewhat less expensive per GB. If you really want a low budget 32TB array, I'd suggest a general purpose storage server (supermicro.com e.g.) rather than a dedicated NAS these days. 67.122.211.84 (talk) 22:09, 15 August 2012 (UTC)

Why doesn't Firefox automatically update on Linux?

Why doesn't Firefox automatically update on Linux? It does on Windows. At work, Firefox has happily updated itself to at least version 11. At home, I have version 4, and even that only because I configured the Linux update manager to be able to download it - otherwise I would still be stuck with version 3. Even the Mozilla Firefox web site says Firefox can be configured to automatically update itself by selecting "Automatically check for updates for Firefox" at the "Updates" tab in the "Advanced" preferences. But there's no such option. There's options for Add-ons and Search engines, but not for Firefox itself. This is in direct contradiction to the screenshots shown on the Mozilla Firefox web page. Does Mozilla have some sort of issue against Linux or is it just a problem with my distribution (Fedora 14 Linux)? JIP | Talk 21:20, 15 August 2012 (UTC)

Linux updates to Firefox usually come from the distribution's update system (locally that's where they should come from, if that's where you got Firefox from in the first place). Fedora 14 is "release no longer supported", so Fedora won't be packaging new versions of anything. -- Finlay McWalterჷTalk 21:29, 15 August 2012 (UTC)

So if I want to update to the newest version of Firefox, do I either have to update the whole operating system, or try to install the latest version of Firefox by hand, and struggle with version conflicts of shared libraries for weeks before I get it to even start up? Or is there an easier way? Updates to Firefox on Windows seem to be coming from Mozilla itself (I can't imagine Microsoft providing them), so why can't this happen for Linux too? JIP | Talk 21:34, 15 August 2012 (UTC)

You should update your OS (Fedora) to the latest stable version, and regularly update it from here on out. You are missing a lot of bug fixes and security patches, just for starters. ¦ Reisio (talk) 21:49, 15 August 2012 (UTC)

Mozilla has no control over third-party Firefox builds on any platform. It's the responsibility of the maker of the custom build to update it, and they don't tend to do that at nearly the rate of Mozilla's official builds. If you install Firefox from mozilla.org you should get automatic updates from Mozilla. There's just one official Linux binary build there; I assume it comes with its own versions of all relevant libraries to avoid .so hell. -- BenRG (talk) 21:43, 15 August 2012 (UTC)

Yeah Mozilla provides Linux binaries, but you'd be a fool to actually use them. :) ¦ Reisio (talk) 21:48, 15 August 2012 (UTC)

Linux systems have package managers, which typically give you the ability to install/remove/update any and all software uniformly, at any time you choose or schedule. There is one manager (per distro), and it works with everything.

Windows does not have a package manager. Each application is responsible for its own update method. There is little to no organization or interaction between them. Each package is responsible for making its own installation system, uninstallation system, update system, and for installing any other packages it needs to work (and checking if they're already installed or not, etc.), etc..
¦ Reisio (talk) 21:46, 15 August 2012 (UTC)

You've chosen a Linux distribution with a deliberately short maintenance window - as the article says "With 6 months between releases, the maintenance period is a very short 13 months for each version". If you need to stay on a release for years, and have it maintained, Fedora is the wrong choice. Ubuntu LTS is supported (meaning you get security patches, not new versions of stuff) for five years; Red Hat Enterprise Linux for longer still (if you pay). -- Finlay McWalterჷTalk 21:58, 15 August 2012 (UTC)

But in general I would say that if your plan is to avoid updating as long as possible, your plan is truly flawed. You should be updating far more frequently than once every 13 months. Most distros have automated update options available these days — use it. ¦ Reisio (talk) 22:03, 15 August 2012 (UTC)

I started using Linux with Red Hat, and switched to Fedora when Red Hat split up to Fedora and Red Hat Enterprise Linux. So I could say I've practically always been using Fedora. Perhaps a switch of distributions would be in order? Ubuntu is still Linux, so I could live with that. I don't want to switch to Microsoft Windows or buy a Mac instead. But would switching to a different distribution require a complete re-install of the entire operating system? Would old binaries from Fedora 14 still work, or would I have to re-install all the programs too? I assume Ubuntu can read EXT4 partitions all OK, so I won't have to reformat my personal data partitions. But really all this seems to be a bit too much just to update Firefox. JIP | Talk 22:05, 15 August 2012 (UTC)

Firefox is the least of your problems; your system, and all its applications, is no longer getting security updates. You have a bunch of vulnerabilities which will never be patched. You shouldn't want to somehow get those old binaries to run on a new system; they're still unpatched. For production systems, where you don't want to update the system every six months, you'd use RHEL, Oracle Linux, Ubuntu LTS, and the like. For a desktop machine of your own, for general purpose use, you want to keep on the latest, or latest but one, release of your dist. There's nothing at all wrong with Fedora, but no-one should be running Fedora 14. -- Finlay McWalterჷTalk 22:11, 15 August 2012 (UTC)

^ What he said. ¦ Reisio (talk) 22:19, 15 August 2012 (UTC)

Google Drive

How do I download the Google Drive app for my PC? The instructions say

Go to http://drive.google.com.

Click the Download Google Drive for your PC button.

but I don't see any such button. Rojomoke (talk) 22:11, 15 August 2012 (UTC)

I see the link on the left hand side pane. What operating system are you using? I'm pretty sure there's no Linux client, so if you are using that to access the site, you may not get the link. - Akamad (talk) 22:32, 15 August 2012 (UTC)

Try https://tools.google.com/dlpage/drive?pli=1#eula. Moondyne (talk) 02:59, 16 August 2012 (UTC)

For me, the link said something like "Free 5GB" and it was in the upper right. Though your mileage may vary. For instance, I already have a Gmail address and a Google Docs account, so I already had the predecessor to GDrive and was signed in to my account. Once I clicked on the button, I was given the chance to download the syncing software for my PC. Dismas|^(talk) 03:10, 16 August 2012 (UTC)

• Information needed: Operating System! You'll not see the download link if you are a Linux user! If you are using Windows or Mac, above information should help! --Tito Dutta ✉ 13:26, 16 August 2012 (UTC)

THe OS is Windows 7, and I do already have a Gooogle account. But logged in or not, I still can't see the links you've given. However, the URL Moondyne gives worked fine. Thanks to all. Rojomoke (talk) 20:02, 16 August 2012 (UTC)

ZFS, heterogeneous storage media and heterogeneous access patterns

I'm looking at buying a solid-state drive to use alongside the hard drive in my Windows 7 gaming workstation, and using FreeNAS or ZFSguru to combine them into a ZFS striped vdev. (I don't need redundant storage.) It occurs to me that the optimal striping pattern will vary between (and sometimes within) files with the frequency of reads versus writes and random versus sequential access, because these affect the two devices' relative overall speeds. Can ZFS monitor these characteristics at the file level, and/or at the byte-range level for gigabyte-plus files, to optimize the striping? Can it predict them for new files based on e.g. data type and folder path? Also, do similar capabilities exist for compression and deduplication? NeonMerlin 23:05, 15 August 2012 (UTC)

That is massive overkill. Just buy a high quality SSD for your system files and other stuff that needs speed, and some regular HD's for your media archive. Make sure to check the 4k random iops in the SSD specs since some not-so-good SSD's are crappy in that important department. 67.122.211.84 (talk) 18:43, 17 August 2012 (UTC)