Wikipedia:Reference desk/Archives/Computing/2010 August 22

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 22

Dual-Domain (Hard Disk Drives)

Hi.

   Why are many modern SAS HDDs dual-ported; what precisely is the meaning and function of dual-domain in this regard, and what are the differences and advantages/disadvantages between dual-domain and single-domain?

   Thank you in advance in all respondents.Rocketshiporion Sunday 8-August-2010, 12:34am GMT

A dual-domain drive can be connected to two host bus controllers (in a coherent fashion, in the same fabric, not just willy nilly). If either dies, traffic can failover to the alternate path through the second HBA. This way no single HBA is a single point of failure - if one dies, the traffic fails over. The operator can decable the drives from the bad HBA, replace it (assuming it's hotplug too, which would only make sense), recable in the replacement, and the system should go back to redundant operation. Fiber Channel already does this. -- Finlay McWalter ☻ Talk 01:52, 22 August 2010 (UTC)

Exact upload date for a file?

Hello. Is there a reliable way to upload a file (of any kind, a document, an executable etc.) somewhere on the Internet, and know the exact date such file was uploaded? The idea is that anyone could prove reliably and easily that the file existed and was uploaded at an exact moment in time. Is there a website that specializes in this sort of thing? Or is there at least some website (file host, etc.) that happens to do this in a clear and visible way? Or is there a way to do this other than with a website, like FTP or something? I have no idea! Sorry if my question sounds dumb... Thanks in advance. -.-; Kreachure (talk) 00:34, 22 August 2010 (UTC)

Only the webmaster of the server to which the file was uploaded could determine the exact date...unless details of the file posting time is given.Smallman12q (talk) 01:46, 22 August 2010 (UTC)

Most FTP services allow the user to see the standard file date information which includes file creation time. It is based on the server's internal clock, which may very well be off my not only minutes, but many years. Therefore, knowing the exact upload date requires checking the file date and knowing how off the server's time was when the file was uploaded. -- kainaw™ 01:48, 22 August 2010 (UTC)

If you are interested in a method to later prove to the world that your document existed at a given date, you may be interested in the ANSI ASC X9.95 Standard for trusted timestamps. -- Tom N (tcncv) talk/contrib 03:27, 22 August 2010 (UTC)

Yup, timestamps were the thing I was looking for. I decided to use a website (Copyclaim) that generates and certifies a timestamp with several hash functions, so that you can confirm the time of submission with either the generated hash or the file itself. Thanks! :) Kreachure (talk) 03:42, 22 August 2010 (UTC)

Copyclaim looks legitimate enough, but they have no visible means of support and might disappear at any time, taking your means of verification with them. A possibly better approach is to make a short text document containing the same information you would send to Copyclaim, and then post that document, or a hash of it, to Usenet. Of course, you could do both. -- BenRG (talk) 07:59, 22 August 2010 (UTC)

We should also make clear that "proving" the original date is different in different communities. To prove authenticity to a computer geek, you can provide technically accurate authenticated timestamps and cryptographic hashes. This is probably the best way to guarantee the content, because it is cryptographically strong. However, proving authenticity in this manner is not pin-for-pin compatible with "proving" in a legal setting. It is more likely that a court would accept a notarized letter or a sworn statement from an attorney, rather than any kind of technical validation of the original date. (This method has flaws, but it will probably be more widely accepted if you have to prove authenticity in court!) The best way to prove original date of creation therefore depends on your purpose. Nimur (talk) 16:07, 22 August 2010 (UTC)

Building a desktop PC - am I missing anything?

Hi everyone. I last built a PC in 2002, so things have changed a bit and I'm not 100% confident I still know what's what anymore. Could you please look at this parts list and let me know if I left out any key components or if something in here is grossly inadequate/inappropriate? The objective is to build a mid/upper-mid range desktop that I can expand upon in the coming years...

• CPU - AMD Phenom II X4 945 3.0GHz 6MB L3 Cache AM3
• Video Card - Sapphire 5770 1G 850MHz 4800MHz 1GB 128-bit GDDR5
• Mobo - GIGABYTE GA-MA770T-UD3P Socket AM3 AMD 770 ATX
• HD - Hitachi 1TB 32MB SATA 3.0Gb/s 3.5
• RAM - 2(or 3) x Kingston DDR3 1333 2GB
• PSU - CoolerMaster GX 750W RS-750-ACAA-E3

I figure network and sound cards are fully on-board these days (yes, it's been that long since I last built a PC!). I was always annoyed by these questions before, but now I find myself forced to ask one. So, thank you for your patience. 61.189.63.176 (talk) 03:51, 22 August 2010 (UTC)

• You will also need a case, a CPU fan, and you will also want some kind of DVD/CD drive/burner. And also the external stuff and software. (mouse monitor keyboard, operating system, power cord) Graeme Bartlett (talk) 04:29, 22 August 2010 (UTC)

I guess that's my punishment for trying to be concise. Yes, I know that I need everything else you listed Graeme. I chose not to include them because they're assumed to be present (to me, at least!) and I didn't want to clutter up my list. Thank you for replying, though. 61.189.63.176 (talk) 04:50, 22 August 2010 (UTC)

I see that you're intending to use the Gigabyte GA-MA770T-UD3P motherboard. Other than the one x16 PCIe 2.0 slot (which you'll be using for your graphics card), this motherboard only has four x1 PCIe 2.0 and two 133MHz PCI slots, both of which are incompatible with the vast majority of present-day expansion cards. If you want a mid-range to high-end desktop which can expand upon in future, you should be looking at a motherboard with atleast; one x16 PCIe 2.0, two x8 PCIe 2.0 and one or two x4 PCIe 2.0 slots. You should also be looking at around 8GB to 12GB of RAM if you intend to run multiple simultaneous programs or memory-intensive programs (e.g. CAD, CAE, gaming, etc.) on the desktop. If you could specify your budget and what you are intending to use this desktop for, I might be able to recommend a more suitable set of components. Are you fixed on AMD, or would you consider Intel? It's less likely that there'll be problems (driver issues, compatibility, etc.) when the motherboard and other components are made by the same manufacturer. As someone who's currently building my own workstation computer, I am quite familiar with the present-day ranges of desktop components. Leave a note on my talk page. Rocketshiporion Sunday 22-August-2010, 6:10am GMT

Can you please enlighten me as to what common accessories uses PCIe x4 and x8 slots?121.74.179.226 (talk) 06:14, 22 August 2010 (UTC)

Realisticly compatibility problems when you buy decent quality components aren't really that common, there's no need to stick with one manufacturer. Nil Einne (talk) 07:00, 22 August 2010 (UTC)

Any reason for choosing Hitachi HDD? 121.74.179.226 (talk) 06:14, 22 August 2010 (UTC)

There's nothing wrong with Hitachi HDDs. I purchased one recently. It's quiet and the price was reasonable (as compared to more prestigious product lines like Western Digital's Caviar Black). People say that they're durable. They used to be branded with the IBM logo, but IBM sold it off to Hitachi a few years ago.--Best Dog Ever (talk) 08:55, 22 August 2010 (UTC)

At least here in Europe, AMD CPUs already come with a heat sink and a fan. 193.236.121.245 (talk) 16:48, 27 August 2010 (UTC)

Internet

My internet comes via a router with an ethernet cable, which is plugged directly into the computer. I have two computers and I used to sometimes transfer the cable to the other computer so it could have the internet, which needed the router to be unplugged and reset otherwise it wouldn't work on the second computer. However in the last week or so when I've done this, the internet has become unusable on both computers for up to three hours. I've tried resetting the router, resetting the computers, running all the troubbleshooting things on Windows which said the problem was "Local Area Connection 2 doesn't have a valid IP configuration". Why is this happening? There is obviously nothing wrong with either computer or the connections, since after three hours it comes back on its own, and it was working perfectly before I transferred the cable, and this method has worked for 2 years until now. 82.44.54.25 (talk) 10:27, 22 August 2010 (UTC)

Did you do anything to your Command prompt? If you release your IP there, you have to renew it. Hazard-SJ ^Talk 10:30, 22 August 2010 (UTC)

I didn't 82.44.54.25 (talk) 10:31, 22 August 2010 (UTC)

What do you think caused the problem? Hazard-SJ ^Talk 10:36, 22 August 2010 (UTC)

If the OP knew, would they be asking? sonia♫ 10:42, 22 August 2010 (UTC)

I said think. Additionally, it could have been caused by something they didn't expect to cause it. It could have also been something they tried, which didn't work well. I remember recently releasing my IP, and I couldn't use the Internet until I renew it. Hazard-SJ ^Talk 10:51, 22 August 2010 (UTC)

To be honest, I'm surprised you had to unplug and reset the router when you moved the cable. Most PCs recognise that a connection has been made through the Ethernet port and request a new IP address from the router using a mechanism known as DHCP. The router would normally respond with an IP address without any other action. I would first check whether you have your PCs set up to use DHCP - ask here if you don't know how to do this, and then I would suspect that the router has not been working properly for a while and is getting worse. --Phil Holmes (talk) 12:32, 22 August 2010 (UTC)

Maybe it's not really a router. --71.141.97.250 (talk) 21:11, 22 August 2010 (UTC)

A router. (This one is wireless, but has regular wired ports, too.)

The other term I've seen occasionally used is "cable modem". Does that make a difference? 82.44.54.25 (talk) 21:24, 22 August 2010 (UTC)

Yes. The "normal" way of networking your computer setup would be to leave the cable modem on all the time, and to connect it to a router, which you would also leave on all the time. You would then connect the router to both computers, so both computers can use the Internet at the same time. Cable modems, like other network equipment, does wear out and die after a while. I assume you didn't purchase the cable modem and that it was provided by your cable company; if I'm right in my assumption, you can just call the cable company, tell them about the problem, and have them give you a new one. I would get a router, too, personally, because of the annoyance of having to fiddle with the plugs, rebooting machines and the cable modem itself ... nuts to that. I see wired routers on sale often for US$20 or so. Comet Tuttle (talk) 20:05, 23 August 2010 (UTC)

They gave me a router but I hate wireless signals so I don't use it. I had a hub but it didn't work, or it wasn't the right thing for connecting the computers. I started a thread on virginmedias forum about it but I called it a router there too, I didn't know it had a special name. Sorry for the confusion. But to clarify, the internet going down for 3 hours when I switch the cable modem from one computer to the other isn't normal? 82.44.54.25 (talk) 20:09, 23 August 2010 (UTC)

Definitely not; there is something wrong. By the way, you can probably use the router in the way I described, while turning off the wireless signal. Plug it all in as above, and follow the instructions that came with the router in order to "administer" the router. Every wireless router I have seen has an option to turn off the wireless radio entirely. (The router administration is usually by using your computer's Web browser to go to an address like 192.168.0.1 and clicking a bunch of buttons in a Web form — your wireless router contains a tiny web server in it to let you administrate it.) If your wireless router also has 4 or 5 wired plugs in the back (like almost every other wireless router in the world) then you're all set — no need to keep juggling the cables, and no wireless signal. Comet Tuttle (talk) 22:36, 23 August 2010 (UTC)

Problems with Windows XP password

After returning from holiday,on switching on my computer,it asked for a password to enter Windows.As it has never done this before,I have absolutely no idea what this is.Trying 'Guest' and blank password tells there is a restriction on and to contact the system administrator.Trying 'Administrator' and a blank password tells me there are restrictions on this account.

Going into safe mode doesn't help as it still asks me for this password before I can do anything,as it does for 'Last Known Good'. The only thing that seems to work is leaving both fields blank-it then comes up with the message 'Loading personal settings..' before very quickly coming up with 'Logging off' and taking me straight back to the password field.

Any ideas as to how to get past this-nothing will come up even with Ctrl-Alt-Delete. The only buttons available are OK,Cancel(which is grayed out),Shut Down and Options.All that happens when you press options is the Shut Down button appears or disappears.

Many thanks... Lemon martini (talk) 14:15, 22 August 2010 (UTC)

You could down download the Ophcrack livecd on another computer and then try to get the password that way. You may have a virus in that you are being logged off as soon as you sign on...Smallman12q (talk) 14:58, 22 August 2010 (UTC)

The error you are getting sounds like something has changed the group policy. It may be best to reformat/reinstall, if you don't have any valuable data on that PC. ^Avic_ennasis @ 16:44, 13 Elul 5770 / 23 August 2010 (UTC)

0phcrack would be useful if you needed the original password, however, in this case, resetting it to a known value would do the trick, and be faster, too. You need a Linux Live CD with chntpw on it. -- 78.43.71.155 (talk) 20:23, 26 August 2010 (UTC)

What utf char ??

In China_Railways_DFH_shunting_locomotives#cite_note-6 the reference contains odd looking latin characters - what are they and how ??? (also I copied them from the title - since it's the reference title should I or should I not change them to standard latin chars ??) Sf5xeplus (talk) 16:14, 22 August 2010 (UTC)

They are fullwidth latin characters in the Halfwidth and Fullwidth Forms section (FFxx) of the Unicode Basic Multilingual Plane. -- Finlay McWalter ☻ Talk 16:44, 22 August 2010 (UTC)

Thank you once again for help with my utf problems.Sf5xeplus (talk) 17:51, 22 August 2010 (UTC)

You're welcome; they're interesting questions that I enjoy figuring out. As to whether you should change the text for the reference, I wouldn't bother - you've accurately characterised the cited reference, there's no need to "improve" on it. -- Finlay McWalter ☻ Talk 18:07, 22 August 2010 (UTC)

ImageMagick help!

I've been searching fruitlessly for a couple of hours now. Can someone please post some code that would resize and convert a whole bunch of JPG files in a folder to 800px width GIFs? Windows XP. FWIW I have IM 6.6.3Q16. As a bonus, how would I right-crop the input images by 120 pixels (equivalently crop the input images to 3336x2304) before processing the above. I'm at the point of tearing my hair out, I just can't find a clear example that indicates what I should do. The help files are also a bit cryptic. Thanks very much, I'll upload the fruits of my labour soon as I can do this process. Zunaid 16:45, 22 August 2010 (UTC)

Your first sentence isn't clear. Do you want to convert (say) 500 JPGs into 500 GIFs, or 500 JPGs into one animated GIF, or 500 JPGs into 50 animated GIFs, or what? -- Finlay McWalter ☻ Talk 16:51, 22 August 2010 (UTC)

500 JPGs into 500GIFs, I have a GIF animator program that will output an animated GIF for me afterwards, but I need to give it GIF files as an input. I'm sure ImageMagick could do this too, but I'm just tired and frustrated at going around in circles through web forums and help files. Zunaid 17:49, 22 August 2010 (UTC)

Resolved

or rather {{fuckit}}. I just found FastStone Photo Resizer which does the job brilliantly. Plus its GUI so much easier to just set up and use forst time. Will post the pic soon as it is ready. Zunaid 18:08, 22 August 2010 (UTC)

The fruits of my labour, behold the mighty Athlone Power Station cooling towers reduced to dust earlier today. Zunaid 18:59, 22 August 2010 (UTC)

It's very good. Remember to put a GFDL tag on it lest the Commons folks delete it. Personally I think the inter-frame delay is about twice what it should be. Also the program you've used has posterised the clouds - does it have a "dither" or "error dither" or "fs dither" option, which should remove those contours we see in the clouds? -- Finlay McWalter ☻ Talk 19:11, 22 August 2010 (UTC)

I hope you don't mind, I reduced the file size to just over 1mb and increased the frame rate a little http://i.imgur.com/OdAay.gif 82.44.54.25 (talk) 19:21, 22 August 2010 (UTC)

The ImageMagick command to take the raw JPGs and emit a downscaled, dithered, animated GIF should look something like this: convert -size 640x480 -delay 40 -dither FloydSteinberg -loop 0 DSC*.jpg athlone_cooling_towers_demolition_2010-08-22.gif      -- Finlay McWalter ☻ Talk 19:18, 22 August 2010 (UTC)

(Marking as unresolved again because we might just improve the current image with people's suggestions.) Finlay, that code doesn't work for me. It crunches for a LONG time but then produced just a 10 byte file. I did change it to -loop 1 and changed the filename filter to match my Canon images. Any other suggestions? Also, I need to crop the right 120 pixels or so, there's a street light on the right of the image that's distracting if left in. I'll also check if FastStone has options for dithering or such-like and see if I can improve it. Zunaid 20:38, 22 August 2010 (UTC)

Could you upload a zip file with the images to rapidshare or some other web hosting site and post the link. From there I'd be able to make the video & animated gif you want...great pictures btw...(its just a hassle to make animated gifs if you don't have the right software.)Smallman12q (talk) 21:33, 22 August 2010 (UTC)

I'm hesitant to do that (also, I've resolved it already, see below). Would you believe that I'm one of probably 10 people to have taken photos from this angle and distance? They detonated the explosion 5 minutes early, just as a heavy rain shower was coming down. Most people around Cape Town got caught with their cameras still in their bags trying to keep them out of the rain. Even local and international broadcasters were caught out. There was a hilarious segment on the live broadcast where the news anchor is still setting the scene, talking to the camera, and suddenly you see the towers exploding behind him. By the time he turned around to look there was only dust left! Oh, and my friend standing next to me recorded it in HD video, I'll speak to him about releasing it. Zunaid 22:06, 22 August 2010 (UTC)

Okay so I solved the issue like so: using Finlay's code as a basis, I first resized and converted each image in place to a separate GIF. Then I used FastStone to crop the right-most pixels (for some reason convert wouldn't crop-and-resize in one step). Then I used UnFREEz to make it into an animation and re-uploaded it (I'm SURE I put the licensing information on the original upload!). So the new and improved animation should be displayed at left shortly. Re-resolved, thanks everyone. Zunaid 21:49, 22 August 2010 (UTC)

OK, but you should know that 6.6MB is kind of a lot for one image, especially once that is of something rather pedestrian (it is not like animations of buildings collapsing, even cooling towers, are that hard to find). You might look into cutting it down in size, i.e. by reducing the color depth. --Mr.98 (talk) 22:20, 22 August 2010 (UTC)

GIF is already 256 colors, can't go much less than that? Also, now it's not thumbnailing. I tried "washing" it through ImageMagick (straight GIF to GIF convert, no parameters whatsoever) which reduced it to 6.16MB, but still no thumbnail. The saga continues...what now? Is this the GIF thumbnailing bug striking again? Re-unresolvedZunaid 22:31, 22 August 2010 (UTC)

128 colors would reduce the size a lot while still keeping relatively good quality 82.44.54.25 (talk) 22:42, 22 August 2010 (UTC)

128 produced. 4.1MB, still no thumbnail. Now what? Zunaid 23:07, 22 August 2010 (UTC)

Thumbnailing is currently broken. At commons, it states There is currently a problem with the creation of thumbnails, and some may not appear at all. Please sit tight, do not remove the images from articles, and hopefully the technical wizards will get it fixed soon.Smallman12q (talk) 01:38, 23 August 2010 (UTC)

search engine query cache

I am having extreme difficulty finding this. I know that all the big search engines cache popular queries. I am looking for some reference that backs the claim. I am not referring to the cache in your web browser. I am referring to caching popular queries on the search engine's server. For example, if 10% of all queries are for "Egg Recall", the search engine will process that query once and then store those results in cache. The following queries will pull from cache instead of searching through all the websites yet again. -- kainaw™ 17:12, 22 August 2010 (UTC)

It's not at all clear what Google caches, and I'd be reluctant to broadly say what "the search engine caches" or doesn't. From what we know about the structure of Google Search (which really isn't that much) it seems to be built on Google Web Server doing MapReduce querying BigTable built on Google File System. The GFS paper (Ghemawat et al.) says it does no data caching. The Wikipedia BigTable article is silent about data caching there too. If Google's front-end web service is structured broadly like Wikipedia's (which we can honestly only guess at) then that leaves three possible locations for caching: caching of complete web pages (which Wikipedia does with Squid), caching of precomputed HTML fragments (which Wikipedia does with memcached) and caching of raw query result [the output of the MapReduce call] (which Wikipedia doesn't cache, and for which I can think of even less reason for Google to cache). We know some parts of Google use memcached (it's part of AppEngine), but not whether Search does. I'm sure you're right, that they do cache a lot of stuff (they'd be mad not to, and they're not mad) but they're darned secretive about exactly where, and exactly what, they do cache. When the Google Dance still occurred (which it seems hasn't been for six or seven years) we could see that different web servers returned searches based on different images of the data (which seemed to be due to incremental replication, rather than caching) but since then (and surely several generations of infrastructure later), while searches are still inconsistent, they're inconsistently inconsistent. So I don't think anyone outside Google knows. -- Finlay McWalter ☻ Talk 18:28, 22 August 2010 (UTC)

I don't have any specific sources about Google but it's a standard technique. You could examine the Lucene code for an example; it's also discussed in the book Lucene in Action and the (old) book Managing Gigabytes. Typically caching doesn't try to track what's "popular". It's just an LRU of some given size. 67.122.209.167 (talk) 07:08, 23 August 2010 (UTC)

Wiki link to odd page

Is their any way I can get this link to work properly as a reference - it seems to be bugging wikipedia..

"ERION Mantenimiento ferroviarro, S.A. , Presentation" (PDF). www.erion.es. Febuary 2008. {{cite web}}: Check date values in: |date= (help) the problem is this I think http://www.erion.es/2%20Presentacin%20de%20ERION%20Feb[1].2008.pdf and the ]

 http://www.erion.es/2%20Presentacin%20de%20ERION%20Feb[1].2008.pdf

..Sf5xeplus (talk) 19:11, 22 August 2010 (UTC)

The brackets are not among the allowed symbols in a wikipedia hyperlink listed here. You need to percent-encode them:

 http://www.erion.es/2%20Presentacin%20de%20ERION%20Feb%5B1%5D.2008.pdf

This works: "ERION Mantenimiento ferroviarro, S.A. , Presentation" (PDF). www.erion.es. Febuary 2008. {{cite web}}: Check date values in: |date= (help)

Thanks, just found the answer myself on http://meta.wikimedia.org/wiki/Help:URL .. using "%5d" and "%5b" for "]" , "["

Resolved

77.86.82.70 (talk) 20:14, 22 August 2010 (UTC)

Matlab - simple column vector formatting question

Hi all,

I've written a Matlab program for a work project which, to cut a long story short, inputs a prime number and outputs a column vector. I need to display the output column vectors one after the other (one for each prime in the range I choose) in a series of columns so they don't take up too much space, but the issue is, the column vectors output vary in length (typically less than 10 elements, but in certain cases it may be more) - normally, I'd simply make a matrix by running the program then adding one column vector after another, but since the length varies I can't simply add one vector after the next in sequential columns, and since there is no 'fixed maximum' for the length of the vectors, and I can't preempt the largest vector in the range of primes I'm looking at in advance, I can't just make a zero matrix with a fixed number of rows and then add in the vectors from the first row down to wherever they happen to stop either.

I've considered writing a loop which changes the number of rows of the matrix every time an output vector arrives which is larger than the current number of rows, before adding the new vector to the matrix, but this seems like a messy option to me, and I'm sure for the more experienced user there's probably a much nicer way to approach the issue. Could anyone suggest anything? What I really want is just a list of column vectors, one after the other, varying in length and preferably without lots of 'zeros' at the end of every vector to fill the space (though any good suggestions would be appreciated) - I tried to draw an ASCII diagram but Wikipedia wasn't happy with the formatting! Thankyou all :) 86.30.204.236 (talk) 22:20, 22 August 2010 (UTC)

If you put a space at the start of the line, MediaWiki will look more kindly on your formatting. The diagram you included looked like this:

k k k k k k  |
k k   k   k  |
k     k      |  PAGE EDGE
k     k      |
      k      |

-------------------  (NEW LINE)

k k k k k k  |
k k k k   k  |
  k k     k  |  PAGE EDGE
  k k        |
             |

(feel free to edit it above so it looks like you intend). -- Finlay McWalter ☻ Talk 22:27, 22 August 2010 (UTC)

Ah thanyou, yes that was about the gist of it! 86.30.204.236 (talk) 23:00, 22 August 2010 (UTC)

There may be a more elegant way to do this, but the following will produce the kind of formatting you are talking about. Obviously, details will need to be tweaked to handle the specifics of your case. The material in angle brackets is meant to be filled in by you.

result = '';

while <NOT DONE CONDITION>
    V = <NEW COLUMN VECTOR>;
    A = num2str(V);
    s = size(A);
    result( 1:s(1), end+1 + (1:s(2)) ) = A;
end

display( result );

Note that if you use this, you will also have to add the new lines by hand if necessary. This snippet doesn't have any internal sense of how many columns to display before generating a new line, but adding that should be straight-foward if desired. Dragons flight (talk) 07:48, 23 August 2010 (UTC)

That's wonderful, I'll give that a go tonight - thank-you very much! 86.30.204.236 (talk) 21:03, 23 August 2010 (UTC)