Talk:Universal Product Code/Archives/2012

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Does anybody understands this page?

Yes -- it's fine. Joe Cetina 03:52, 5 Mar 2004 (UTC)


I missunderstod the following

  1. Add the odd-numbered digits (first, third, fifth, etc.) together and multiply by three.
  2. Add the even-numbered digits (second, fourth, sixth, etc.) to the result.

Read it as "Add the odd digits" and "Add the even digits", maybe some other wording would be helpfull?

ex:

  1. Add the digits on odd positions (first, third, fifth, etc.) together and multiply by three.
  2. Add the digits on even positions (second, fourth, sixth, etc.) to the result.

ok--Gbleem 17:06, 14 December 2005 (UTC)


checksum algorithm

"Details on the checksum algorithm" is broken. Maybe for the author it will be easier to find it again?

I thought the standard for the barcode is different than the Universal Product Code. --Gbleem 17:06, 14 December 2005 (UTC)


The algorithm for checksum shown only works if the first digit is 0. It should read add the odd digits 3rd+5th+7th etc...it should not include the 1st digit (the identifier)


The algorithm description is semantically incorrect, but it does work. Strictly speaking, the digits are numbered from the right, rather than the left. The checksum is digit number 1 and the number system identifier is digit number 12. The sum of the odd numbered digits, plus three times the sum of the even numbered digits must come to a multiple of 10. The specification numbers the digits from the right, because the same kind of algorithm is used in several symbologies with varying numbers of digits, and numbering from the right ensures that the last digit (the checksum) is weight with 1 rather than 3 irrespective of the number of digits in the code. The number system digit (the first digit on the left) SHOULD be included in the checksum calculation David ORourke (talk) 13:47, 18 September 2008 (UTC)

Having found a copy of the current specification, this needs some further thought. previously digits were numbered from right to left. But this is no longer the case. Nevertheless, the check digit always has a weight of 1, so the weights for the odd or even numbered digits depend on the total number of digits in the code. Since this article is specifically about the UPC and seems to refer in main to the UPC-A which always has 12 digits. It's probably best not to confuse the article with alternatives. I still think that it needs some clarification, as people are obviously confused by the current explanation. The official specification can currently be found in this pdf David ORourke (talk) 16:10, 22 September 2008 (UTC)

And EAN?

"In addition, this also expands the numbers available for the U.S. and Canada by 50%, adding 10 to 14 to the 00 to 09 (0 to 9 in UPC) already in use." If they reserved 0 for upc then why 11, 12 and 13? Am I misreading this?? --Gbleem 17:15, 14 December 2005 (UTC)

See the Country Codes section below. --[REL] 2007 July 2

what about scaling, cropping height and 8 digit versions?

--Gbleem 17:15, 14 December 2005 (UTC)

Article says the nominal height is 25.9mm = 1.0 in. Of course 25.4mm is 1.0 in; 25.9mm is ~1.02in. Presumably the 1in measurement is the correct one? —Preceding unsigned comment added by 121.73.163.79 (talk) 06:53, 20 March 2009 (UTC)

Universal or Uniform??

"Universal" per the excerpt of some book I have but can't find the title. Someone else, hopefully, will list a reference. --[REL] 2007 July 2

--Gbleem 18:45, 14 December 2005 (UTC)


It is Universal. I think the confusion arises because the body which issues numbers in the United States was previously the Uniform Code Council. This is now the responsibility of GS1. David ORourke (talk) 13:50, 18 September 2008 (UTC)

Email from George

--Gbleem 05:51, 16 December 2005 (UTC) I sent an email to George J. Laurer asking him to look our article and this is his reply:

Dear Mr. Martin,

I hate to bust your bubble, but I believe you have numerous errors in your articles and the structure of the articles certainly misleads readers.

For the most part this comes from the references you cite which in themselves are not correct.


Article “Universal Product Code” As an example: You state “Joseph Woodland and Bernard Silver patented a bullseye[sic] style code in 1952 and the first commercial [sic] use of barcodes was in 1966.” Their patent was filed in 1949. You infer that it was used in 1966 which it was not. You say “In 1970 Logicon Inc. created the Universal Grocery Products Identification Code (UGPIC). In 1970 it was used by Monarch Marking in the United States and Plessey Telecommunications in the United Kingdom.[3] (a consulting firm) in conjunction with UGPCC”. Again not true at all. Read Revolution at the Checkout Counter by Stephen A. Brown. He was the council for UGPIC.

Your section “Current Code” makes statements which are then contradicted in the section “Representation”. An example is: “Company prefixes are assigned by EAN-UCC, which is now using longer company codes (with shorter item codes) for smaller companies.”

The section “Expansion” is entirely incorrect.

Further, UPC should not be used. The correct abbreviation is U.P.C. with periods.

Article “Barcode” There are many errors and omissions. I suggest checking the book Punched Cards to Bar Codes by Benjamin Nelson.

Jeff, this reminds me of the game we played in grade school. One person would tell the person next to him a short story. That person told the next person and so on around the room. When it got back to the originator, it bore no relation to the original story. I believe that is how the inaccuracies crept into your article. Many of your references contain the same inaccuracies.

I think your articles are important and needed and I commend you for your effort. However in their present form they to more harm than good. More research is needed. I wish I had the time to rewrite your copy.

Bad example?

For example, let's say the the first digit in a barcode, after the 1-1-1 start code, is one

what 1-1-1 start code? I only know of the mentioned 101 code. --Abdull 19:43, 29 December 2005 (UTC)


Number of unique items

How many unique items can be indexed with UPC / EAN? --Abdull 19:43, 29 December 2005 (UTC)

Country Codes

There's a section for Country Codes in the article, that starts "The first 3 digits of the bar code of any product you buy represents the country." However, the rest of the article states that a UPC code is twelve digits, starting with one digit for product type and five for manufacturer code. Nowhere else in the article mentions country codes. So why mention these numbers? 12 != 3+12 --Mozai 19:22, 17 April 2006 (UTC)

When you consider the U.P.C.-A the same as the EAN-13 if you right justify a 13-digit number (including the check digit) and look at the country codes, then it makes sense. The bar codes for U.P.C.-A and EAN-13 are the same--but the listing of the numbers on the packaging isn't. --[REL] 2007 July 2


The bar codes for U.P.C.-A and EAN-13 are not the same. Except where the first digit of the EAN is 0. Since this article is about UPC, I think that the references to three digit country codes should not be here. David ORourke (talk) 13:53, 18 September 2008 (UTC)

Example image is wrong

I was implementing code based on this page and ran into trouble -- which I eventually realized was my own fault, the result of my example image being an EAN code, not a UPC code -- but part of the problem was the example image on this page is also EAN, not UPC. Look at the code for the digit '2' -- it uses the EAN odd parity code 2122, not the code 2212 from the table in the representation page.

Forget that. So far as I can tell, the representation section is wrong. It mixes even-parity and odd-parity UPC codes. I will edit it to be consistent with other information on UPC on the net and the example image.

UPC-A Is Used As An Undefined Term

The page uses UPC-A as an undefined term. It starts talking about UPC-A without giving an overview that says what the relation of UPC-A is to UPC or EAN. Is there a UPC-B? If so does it have a different checksum algorithm? Or is UPC-A simply a synonym for UPC? It would be nice if someone could fix these things on this page. In general the term "UPC-A" should be defined before it is used. - ATBS 14Jun06

There is a U.P.C.-E which is a condensed version of the U.P.C.-A. --[REL] 2007 July 2

Dividers

What digit are the double, long, thin, bars that divide the code into sections, and why are they there? Is it 6, as said by a character in the movie Naked.83.118.38.37 23:18, 28 August 2006 (UTC)

Length of Company Prefix not simply "6"

According to the GS1 US FAQ[1], the "U.P.C. Company Prefixes can vary from six to ten digits in length".

"A further stimulus is the fact that in 2005, associated with the 2005 Sunrise program, GS1 US has

begun issuing variable-length GS1 Company Prefixes and retailers are expected to accept imported products identified with GS1 Company Prefixes. Both changes will lead to an increasing number of coupon mis-redemptions if the full Company Prefix is not processed. This will impact retailers, manufacturers, and coupon processing agents." taken from "North American Coupon Application Guideline using GS1 DataBar (RSS) Expanded® Symbols" (27 June 2007) written by GS1 US--[REL] 2007 July 23

History Section

there's a lot of history of UPCs in the barcode article. Should that be moved here?


I moved my "Development of the IBM UPC proposal".UPCMaker 21:50, 14 August 2007 (UTC)

Periods in U.P.C.

I have added periods in most of the instances of UPC in this article. (the main ones I didn't were inside of links) in accordance with George's letter. I have been reading quite a bit about barcodes over the past year or so, and it is supposed to be U.P.C. (I'm pretty sure the reason is legal, to differentiate from Uniform Plumbing Code.) VikÞor | Talk 05:02, 26 September 2006 (UTC)

No periods should be used. Standard usage is without, and as an acronym no periods is standard. Regardless of whether "George" likes them there or not, actual usage is definitely without periods. 172.151.113.138 23:27, 1 November 2006 (UTC)
People say, "ain't" but as an editor writing a formal document, I'm going to write, "is not" instead. If a court of law made the distinction, I'll make the distinction. --[REL] 2007 July 2

UPC Wiki

I added an external link to www.upcwiki.com. In the interest of full disclosure, I am the owner of that site. UPC Wiki is an attempt to create a complete product catalog of all retail goods in the world, identified by their unique barcode. In any case, if you have any questions or concerns with my addition of the link, please let me know. But I feel that the site and the addition of the link provides value to Wikipedia. If you feel otherwise, please provide constructive feedback on how I may improve things.

kind regards,

--Casey Plummer 15:22, 12 November 2006 (UTC)

UPC-A

Isn't UPC-A S-5-M-5-Check (SLLLLLMRRRRRX), not S-6-M-6-Check as stated in the article? I know all the products with UPC-A bars within easy reach of me are this way; Also EAN seems to be S-6-M-5-Check (Read "EAN and UPC are the same Symbol" near the end of this article) --tonsofpcs (Talk) 17:09, 24 December 2006 (UTC)

UPC-A is S-N-LLLLL-M-RRRRR-X-S There is a start guard pattern, a number system identifier, a 5 digit manufacturer code, a Middle guard pattern, a 5 digit product code, a checksum and a stop guard pattern. The start and stop patterns are identical. Note that since the original design of the code, there are now longer manufacturer codes available, with correspondingly shorter product codes, but the middle guard pattern is always in the center of the barcode. EAN-13 is not the same symbol unless the EAN begins with 0. It has an additional Character near the beginning. S-C-N-LLLLL-M-RRRRR-X-S David ORourke (talk) 13:59, 18 September 2008 (UTC)

UPC Image

UPC Image is wrong - according to GS1 General specifications first and last digit of UPC-A should be printed outside the symbol to indicate boundaries of the quiet zones (which are necessary for th e bar-code scanner to scan properly. I will be submitting the right symbol representation shortly. Gs1mo 12:54, 14 May 2007 (UTC)

Redundant "code"

There are several instances of the redundant usage of "code" after UPC. Shouldn't these be removed? Check out the article on "RAS syndrome".

..."ATM machine," yes, it's redundant.
x favor ayudenme necesito ls codigos de barra de los paises mandenlo a gatita_salvaje_807@hotmail.com

Trimming the External links section

At present most of this article's external links are not compliant with our Wikipedia:External links guideline. There are one or two links that provide historical info about barcodes. I'd be tempted to keep these, since getting reliable published info has not been easy. Most of the rest are promoting commercial products or services, and these should go. (Wikipedia is not a business directory). Please give your opinion if you think any of these should be kept. EdJohnston (talk) 22:55, 13 December 2007 (UTC)

ISBN mapping?

ISBNs are mapped to EAN (and vice versa) with the Bookland code. Is there any such mapping with UPC? A look at my book shelf did not reveal any obvious correlation. Thanks, Tim Landscheidt (talk) 23:45, 23 January 2008 (UTC)

There is no such mapping to UPC-A. As I understand it UPC-A codes are officially deprecated, and have been for a long time. If you just append a 0 to the beginning, it maps to it's corresponding EAN-13. The bookland codes all begin with 978, and follow simply with the 10 digit ISBN, with its checksum removed, and a new EAN checksum.David ORourke (talk) 14:03, 18 September 2008 (UTC)
Thanks, but my interest lay more with mapping UPCs (from my barcode scanner) to ISBNs (from my books database). In the end, I found that not only there seems to be no easy mapping, the UPC is also not unique without the +5 addon – several books on my shelves share the same twelve first digits :-( (BTW, some googling found no indication (and one statement to the contrary) that the add-ons are checksummed in any way). --Tim Landscheidt (talk) 19:55, 31 July 2009 (UTC)

666 Barcodes?

Worth mentioning that the end and middle bars "look" like one of the sixes? Hence some believe that there is '666' (the number of the beast) on every barcode. Pretty amusing. Might get deleted. Aronzak (talk) 13:37, 29 July 2008 (UTC)

See Elephant in the room. __meco (talk) 17:26, 25 May 2009 (UTC)

Optically Inverted

Before the Middle guard bars, a 1 indicates a bar, while a 0 indicates a space. After the Middle guard bars, however, the patterns are optically inverted. In other words, a 1 now indicates a space, and a 0 now indicates a bar. In the illustration above, the "4" digit (shown in detail), falls after the Middle guard bars, causing the pattern of bars and spaces to be inverted.

From the diagram on the page and a barcode I have in front of me, it appears that the end bit sequence 101 is not included in the optical inversion, just the last six payload digits. If this is correct, a clarification is in order. Workaphobia (talk) 23:16, 6 September 2008 (UTC)

This is indeed the case, the guard pattern is not part of the 'optically inverted' part of the code. I agree that clarification is needed. David ORourke (talk) 14:04, 18 September 2008 (UTC)

Talk page banners

Just an FYI: An IP user has recently added a number of banners to this talk page. I have removed {{reqphoto}} pending clarification as to just what sort of photo is requested. I have also requested review from the WikiProjects whose banners were added, to verify that this article really is within their scope. Anomie 21:27, 13 September 2008 (UTC)

Current "Circular" UPC Codes

- No mention of the "3-ring bullseye centered in a square field of dots", as used by courier companies (UPS). Wapiti (talk) 17:03, 12 December 2008 (UTC)

That is because a MaxiCode is not a Universal Product Code. Anomie 17:47, 12 December 2008 (UTC)
- Yes, that's the name, and yes, it's not a UPC, and yes, I missed the reference to "bar code" in the first line of the article probably because the first instance is not linked (which I will now fix). Thank you very much!
      • SECTION "Current 'Circular' UPC Codes" CLOSED [hopefully] Wapiti (talk) 19:44, 12 December 2008 (UTC)

Who issues UPC codes?

- There doesn't seem to be anything about who issues or controls the issuance of UPC codes. Does anyone know? Google seems to point toward "GS1" as the official source. Seems like this might be worth a paragraph or two. —Preceding unsigned comment added by 206.126.80.142 (talk) 20:59, 20 December 2008 (UTC)

Perhaps this should be linked to the Spanish one: http://es.wikipedia.org/wiki/C%C3%B3digo_de_barras —Preceding unsigned comment added by 140.147.236.194 (talk) 14:21, 5 June 2009 (UTC)

There's an organization in each country to which GS1 delegates control. In the US, it's to a company now known as GS1 US. There are many resellers in the US.--Elvey (talk) 05:30, 1 December 2009 (UTC)

New UPCs...

Lately I've seen UPCs that have 3 different bar codes.....1 long one, & 2 shorter ones...usually on coupons. What's the story?

67.83.181.248 (talk) 01:54, 2 September 2009 (UTC)cyan

The shorter barcodes encode expiry date and sometimes the "offer/reward" how much you have to buy to get the discount which may be a product, a percentage off, or fixed money off. Interestingly, the EAN/GTIN equivalent for coupons (98, 99 prefixes) don't allow for offer/reward complexity or for supplemental barcodes, they just embed manufacturer prefix and money off (with an implied currency in 98 prefix to e. g. handle Euro conversion for countries in transition to the Euro). — Preceding unsigned comment added by 78.105.60.169 (talk) 12:15, 17 September 2011 (UTC)

Some online Sources/References

--Elvey (talk) 23:37, 8 October 2009 (UTC)

Optically Inverted? Right-to-Left?

The explanation of bit representations is confusing or wrong. For example #4 is L:0100011 R1011100 but the following paragraph states:

"Before the Middle guard bars, a binary 1 is indicated by a bar, while a 0 is indicated by a space. After the Middle guard bars, however, the patterns are optically inverted."

The pictured representation of 4 is BAR-SPACE-BAR-BAR-BAR-SPACE-SPACE, which matches the table, and is clearly AFTER the middle? Unless you're reading right-to-left? That can't be, since:

"In the illustration above, the "4" digit (shown in detail), falls after the Middle guard bars, causing the pattern of bars and spaces to be inverted."

Somebody please clarify. (preferably in the article too) --70.128.123.6 (talk) 20:34, 9 October 2009 (UTC)

The Universal Product Code (UPC) is not a barcode

The Universal Product Code is not a barcode. It is a number that is assigned to uniquely identify one product. Because the number is often encoded on products by a barcode so that it can be read by a scanner, some confusion has arisen. The formats for UPC's vary depending on the type of the UPC.

For example, the format for an item (or each) UPC is generally denoted by (1-5-5-1). That means that it has the format p mmmmm nnnnn c, where the p is a prefix that denotes the industry (0, 6, or 7 denoted general merchandise, 3 denoted drug), mmmmm is the manufacturer's id, nnnnn is the item number, and the c is a check digit. The format for a case UPC is generally denoted by (2-5-5) or pp mmmmm nnnnn, where the p,m and n stand for the same things they do on the item UPC. For the case UPC, the prefix can be 00, 06 or 07 for general merchandise, or a 03 for drug. You will also notice that the case UPC lacks a check digit.

The information on the formats of the UPC's was originally available from the Uniform Code Council. I don't know if you can still find the information on GS1 or not.Martyedi (talk) 23:13, 22 October 2009 (UTC)

Do you have a citation to support that? A UPC code is made up of 1)human-readable digits, and 2)those same digits, encoded with (a modified version of) the Delta C Code. See the historical documents that are used as sources. Without the barcode (or a similar technology) unique product numbers would be close to useless, and hence quite possibly wouldn't exist. Early on, there was a Uniform Grocery Product Code Council, which became Uniform Product Code Council, Inc, which became GS1, and in-between, was the Uniform Code Council. There are lots of companies that sell legitimate U.P.C. numbers <sic>, not just GS1, much like there are many legitimate (ICANN-accredited) domain name registrars, not just Network Solutions. That doesn't mean the info in your second paragraph isn't welcome, especially if you can WP:RS source it, though most of it is already here.--Elvey (talk) 03:47, 1 December 2009 (UTC)

UPC-E

I restored the UPC-E information. They're real; I'm holding one of these: http://www.upcdatabase.com/item/04206828 and a google search shows they're popular.

Produce Electronic Identification Board (PEIB)

This page needs a description (and a link out to a stub page) of the UPC codes assigned by the Produce Electronic Identification Board (PEIB), who maintain a standard 4/5 digit code scheme for fruit and veg internationally. These numbers are often seen on peel-off labels on fruit, especially apples, to assist checkout at retail and in the supply chain. They are also often used as the product code in a price or weight embedded barcode produced when product is weighed/packaged in store.

Example 4101 Braeburn Apple

PEIB has a UPC manufacturer prefix (033383, but not often used) for prepacking of product sold by the each--the latter part of the UPC will be the 4/5 digit PEIB code.

further details http://www.plucodes.com/faqs.aspx

Full list can be downloaded by http://www.plucodes.com/search_wizard.aspx?s=1 (you will need to enter country, role, company, then View All)

http://www.google.co.uk/search? hl=en&safe=off&rlz=1T4ADSA_enGB341GB343&num=100&q=Produce+Electronic+Identification+Board+list&meta=&aq=f&oq=

94.125.16.11 (talk) 12:21, 24 December 2009 (UTC) Milest (anonymous)

Errors in reading

When I scanned with "upcode" the UPC-A barcode number, it resulted in 0981654327098, instead of 987654321098. Why happens so? — Preceding unsigned comment added by 46.12.93.46 (talk) 10:53, 29 May 2011 (UTC)

Incorrect Images for UPC Encoding (Number 2)

The images for the UPC Encoding for the number 2 are incorrect (both left and right). The image is showing 0111101 instead of 0010011 as is listed below the image. The images appear to be the same as for the number 3.66.179.186.164 (talk) 22:10, 6 July 2011 (UTC)

Conspiracy Theories

Hi! This type of barcode is the most seen on consumer goods here in Greece. Some friends of mine suggest that the barcode is the mark of the Beast and whoever accepts it surrenders his soul to the Anti-Christ and will go to hell. We weren't taught Revelation at school so I don't really know what the Beast thing is and why it needs the barcode so I can't confirm weather my friend's views of Revelation and barcodes are accurate.

BUT the UPC does have the tree pairs of double lines as scanning guides, which are the symbols of 6.

Who and why would choose to use as guide marks the number of six and not some other digit to avoid this problem? — Preceding unsigned comment added by Galanom (talkcontribs) 01:10, 24 August 2011 (UTC)

bad math

last bullet point:

  • "In North America the EAN adds 40% more codes mainly by adding 10 to 13 to the 00 to 09 (0 to 9 in UPC) already in use. This is a powerful incentive to phase out the UPC."

I believe that this is an exponential function, 40% being accurate only for a 1 digit upc. On a 12 digit 00-09 code we'd have 10^12 possibilities: 1,000,000,000,000. Adding 4 more digits makes this 14^12: 56,693,912,375,296. not 40% more codes... 5600% ... and it gets exponentially larger for larger than 12 digit codes. I'm not sure how this should be worded to be clear to read and obvious. -- Vaevictus (talk) 15:28, 7 August 2008 (UTC)

The math in the page is correct, yours is wrong. The total numberspace for EAN codes is 1012 (1,000,000,000,000). Codes beginning 00-09 comprise only 10 * 1010 (100,000,000,000) of these. Codes beginning 00-13 comprise 14 * 1010 (140,000,000,000), which is 40% more. 5669.4% would be correct if the change was from base 10 to base 14 numbering rather than simply allocating an extra 4 prefixes for US use. Anomie 16:03, 7 August 2008 (UTC)

I disagree with this: "UPC-E: 10 possible values per digit ^ 2 possible parities per digit ^ 6 digits = 100 permutations per digits ^ 6 digits = 1,000,000,000,000" Shouldn't it be "(10 possible values per digit * 2 possible parities per digit) ^ 6 digits = 20 permutations per digit ^ 6 digits = 64,000,000"? — Preceding unsigned comment added by 68.3.200.113 (talk) 14:39, 2 May 2012 (UTC)

Extremely confusing

I am a layman and know nothing about coding. I find this article to be very technical and to assume a lot of knowledge and context about bits, coding, and so on. The first paragraph, for example, is incomprehensible to me, the illustration raises more question than it answers, and the table is not properly introduced or explained. There are references to "timing" and "width" (of what?), and references to L and R as "digits", and an explanation of "prefixes" that has no introduction or definitions, and a section on "interpreting" that doesn't actually tell me how the codes are read and interpreted. (This is exacerbated by my personal pet peeve, which I think is shared by many people, when technical articles make you click on every blue wiki-link to get basic definitions of the crucial terms and references in the article).

It would be great if someone could add a paragraph to the lede and expand the sections below to make this article more engaging to someone who doesn't already know a lot about coding. 71.237.216.198 (talk) 13:01, 16 June 2011 (UTC)

I fully agree. The article tries to explain the algorithm behind UPC but is too difficult to understand. --Shandristhe azylean 18:05, 12 November 2011 (UTC)

You have the option of tagging the article with the template: {{technical}} -- œ 08:13, 20 June 2012 (UTC)

How exactly the weight or price is encoded

If the first R will determine how...

But I can't find anywhere the exact information what it'll be and how it tell be is it weight or price? — Preceding unsigned comment added by MWaltari (talkcontribs) 17:50, 21 June 2012 (UTC)

Look at the 1st question here...

If you look at the 1st question here from 2005, and similar complaints up to June 2012, and everything is said. Never saw such a bad, confusing article in Wikipedia. If somebody knows something about the code, he should edit the article, it cannot get worse. Also the images are confusing, a) to big, with numbers below that do not match whatever (so why those numbers?), and b) e.g. about the last images (IBM) no explanation why there are red crosses at distinct positions. OMG! 93.134.176.123 (talk) 11:21, 15 July 2012 (UTC)

super-imposed numbers

In the standard UPC-A illustration the numbers 789999 are superimposed on 123456

Yet if you do a copy and paste you get UPC-A Q.svg 123456 789999.

Is this deliberate? I have never seen numbers superimposed on each other in real life.

81.154.150.89 (talk) 21:18, 5 September 2012 (UTC)