Talk:International Bank Account Number/Archive 1

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Archive 1 Archive 2


The National Bank of Kazahkstan has recently announced the use of IBANs. The details of the Kazahk IBAN are taken from various sources. Once this IBAN is in the official IBAN Registry, these links and this comment should be removed. Martinvl (talk) 14:09, 18 March 2010 (UTC)

Kazakhstan has been in the IBAN Registry since version 20, September 2010. BBAN structure is given as 3n,13c - not 3n,3c,10n as it still is in Wikipedia. — Preceding unsigned comment added by (talk) 07:29, 5 June 2012 (UTC)

security advice

I see the statement under security and privacy:

"Since an IBAN is nothing more than an enhanced bank account number, account holders should give the same security considerations to their IBANs as they do to their bank account numbers."

Presumeably, wikipedia should not be giving advice.

Moreover, I think this is poor advice. An ACH account number in the US can be used to print phony checks, or draw money out of electronically. Whereas it's quite common to print IBAN numbers on business cards, letterheads, and other stationary, because the accounts are implemented in a way that criminals cannot externally draw from them. Jgombos (talk) 11:01, 14 February 2010 (UTC)


This page is for code junkies not for anyone looking for any useful link to a service that actually gives you a human manageable way of calculating a result from input data �Preceding unsigned comment added by (talk) 14:34, 23 May 2008 (UTC)

in the examples its totally unclear what those kk and CCCC and so on mean

I agree. That needs fixing.  :) — Helpful Dave 13:04, 9 Apr 2005 (UTC)
See my comments below. My spreadsheet is used by the OECd amongst others - OR beware. —Preceding unsigned comment added by (talk) 10:22, 17 August 2010 (UTC)

Integrate Romanian national standard

The link Regulation regarding the usage of the IBAN codes in Romania explains in detail the IBAN codes in Romania: format, usage etc. It should really be added as it contains useful information in sync with the content of the article.

Everything in the Romanian document that is of interest to IBANs as a whole is already in the Wikipedia article. The document might be an appropriate reference to an article on law or banking in Romania. Martinvl (talk) 12:23, 29 July 2010 (UTC)

countries, where IBAN is applied?

On the official page there are some lists. Also there is a sub-list of IBAN applicants that also apply some TR201 regulations and SWIFT methods...

But I have a brochure from our central bank about the imminent adoption of IBAN here at home. In the brochure IBAN is described as "Until Jan2006 IBAN is applied in 42 states, including all members of the EU". The lists on the official page seem shorter (if we count all entries - states and overseas territories - there are more than 42. If we count states only - there are less than 42). (talk • contribs • email)

[According to the dropdown lists at IBAN seems to be used in all of the EU including all dependent territories, plus the following countries and dependent territories of EU countries (including dependent territories that aren't part of the EU):] Delete, this is a 404.

Andorra Bosnia and Herzegovina Croatia Iceland Liechtenstein Macedonia, Former Yugoslav Republic of Monaco (identical to the French format except for the country code 'MC') Norway Romania San Marino (identical to the Italian format except for the country code 'SM') Serbia and Montenegro Switzerland Tunisia Turkey United Kingdom

Not all of the above countries are mentioned on the IBAN page. (Change: see the new SWIFT link for an up-to-date list)

Stefan2 04:55, 15 May 2006 (UTC)

The IBAN system is open to all countries, and it is usual for the country to prepare for its adoption before formally announcing the fact, usually because it requires the account number structure of all bank accounts in the country to be standardized such that no account is ever duplicated. That may require extensive coordination between banks before plans can then be made to convert the accounts yet again into the IBAN format, and very often the banks in small countries lack the formal relationships with Government necessary to do so legally: this was certainly the case in Malta, which I was involved in progressing.
Also, beware that the BIC and SWIFT codes are NOT guaranteed to be identical under all circumstances: always check with the beneficiary of in doubt. —Preceding unsigned comment added by (talk) 09:53, 17 August 2010 (UTC)

Swedish IBANs

It says:

   * Swedish IBAN format: SEkk BBBB CCCC CCCC CCCC CCCC
       The Bs represent the bank code and the Cs the account number.

This needs some explanation. What is the bank code? At first I thought it was the sort code ("clearingnummer"), but then I realised that the BBBB part in my IBAN is 3000, while my sort code is 3300. Secondly, many of FöreningsSparbanken's sort codes are 5-digit codes instead of 4-digit as used by most other banks, and postal giro accounts, while they have IBANs, don't have sort codes at all.

Stefan2 04:25, 15 May 2006 (UTC)

There are two schools of thought inside the governmental communities concerning the amount of detail needed in its part of the IBAN field. Those countries with long-established standardisation such as Belgium go straight for the account number, relying on the BIC element to cross-check it. Others like the UK prefer to prefix the number with a repetition, often in 4 characters such as LOYD for Lloyds TSB. This is probably the case in Sweden as well, but please be aware that this is not the place to discuss personal needs: you should ask your bank to explain. —Preceding unsigned comment added by (talk) 09:58, 17 August 2010 (UTC)

Please verify.....

The first paragraph on the front page:

"....and was later adopted as ISO 13616:1997 and now as ISO 13616:2003"

The info provided by this sentence doesn't seem to be correct according to the webpage at

To my knowledge, if a standard is published and later revised, the history of the versions will be listed. See this page for more information

�The preceding unsigned comment was added by (talk) 12:12, 16 April 2007 (UTC).

Why is Morocco listed in the IBAN examples ? Any source for this information ?

�The preceding unsigned comment was added by (talk) 14:41, August 20, 2007 (UTC)

There seems to be an issue with the checksum calculation. The English version of this page specifies minus one, modulo 97 and plus one on the remainder again. This is to avoid a zero value.

However, many other sources ommit this minus-one-plus-one operation, and simply specify modulo 97. Notable examples are the description on the Dutch version of this page [[2]] and the algorithm implemented on the German site for IBAN calculations [[3]]. It is not clear to me what the ISO specifications say on this issue, as they cannot easily be consulted.

As an example consider the German domestic account number BLZ 50040000 KTO 0336665500, for which both variants yield a different checksum, 01 versus 98. The correct IBAN seems to be DE98500400000336665500.

Is this a genuine example? Both variants will be passed by the IBAN validator as being valid. If it is genuine, a note to this effect might be useful. --Martinvl (talk) 14:18, 18 June 2009 (UTC)

I have removed the minus-one-plus-one operation to bring the text into line with the text of numerous other sources. The zero result for the mod operation is of no consequence since this will yield a checksum of 98 and 98 mod 97 = 1. May I also draw to attention that there is a warning in certain sources that only banks should perform the actual IBAN calculation. Martinvl (talk) 12:06, 3 July 2009 (UTC)

It is more complicated than that, because the detail contains characters, in the National ID in the first 2 positions and often in the nationally-defined fields. Firstly, the first 4 characters are cut from the front and added to the end of the string, substituting 00 for the as yet unknown checkdigits in 3rd and 4th position. Then, any letter is substituted with 2 digits (A=10, B=11...Z=35 so ABNA becomes 10112310). Finally, a Mod 97 calculation is performed, and the result is deducted from 98 to give the check digits. As many PCs can't handle that long a number, do it iteratively on the first 8 digits, tacking the MOD 97 result onto the next 6 and so on until you get your answer. —Preceding unsigned comment added by (talk) 10:14, 17 August 2010 (UTC)

ISO 7064 mod 97-10

What is "ISO 7064 mod 97-10"? I am reading the document þ EBS204_V3.2] and after saying that ISO 7064 mod 97-10 should be applied, it is said that should be 1 the remainder of the division of the number by 97. So this is just that the whole number should be congruent to 1 mod 97. There no need to invoke an ISO standard (and what id ISO 7064) and why there is a 10 after the 97? In the standard and here is written a sentence that is much more complicated than what is required. -- AnyFile (talk) 20:16, 28 May 2008 (UTC)

See my full explanation above. —Preceding unsigned comment added by (talk) 10:21, 17 August 2010 (UTC)

Moldova has IBAN too

Sorry for barging in. I am in Moldova and I just received a new bank card and my account has an IBAN code: MD11 MOBB nnnl lnnn nnnn nnnn (the first two groups are as is, i guess country and bank, the rest: n is a digit, l is a letter). Hope this helps (talk) 10:48, 9 October 2008 (UTC)Vitalie

The supposed Moldovan 'MD' prefix is not listed at the official IBAN page. prat (talk) 11:51, 29 July 2010 (UTC)
It doesn't matter as long as it's the ISO Code for Moldova. It's worth keeping an eye on that ISO page as if for any reason a country's ISO code changes, so will the bank account of everyone in the country (see my checkdigit explanation above). —Preceding unsigned comment added by (talk) 10:16, 17 August 2010 (UTC)

Similar Domestic Systems

The article has a section called "similar domestic systems".

This is rather stupid. The whole point of IBAN is that it is INTERNATIONAL and not NATIONAL! We could end up listing the national bank routing systems for hundreds of countries but this will end up telling us NOTHING about IBAN. As far as I know IBAN is the only attempt at providing a transnational routing code for a bank account that can work globally.

For this reason I propose that we delete this section. It tells us nothing about IBAN and is a distraction.--Tom (talk) 19:25, 10 November 2008 (UTC)

The European Union will soon dispense with National account number structures under the SEPA Directive, all banking will use BIC/IBAN. This is why.

Is this pure math?

I'm a bit disgusted with Hauskalainen's censorship of information on the dangers/pitfalls of IBAN wire transfers. Does everyone agree that discussion about a number should strictly be confined to the math, and not the use of the number? If so, the responsible action is to move the content, not destroy it by vandalizing the security and privacy section. Jgombos (talk) 12:33, 20 June 2009 (UTC)

I agree with Jgombos. However the security and privicy section read a little like an anti-advertisement. It should have been rewritten emphasising that IBAN numbers was an addition to the current banking system, not an alternative. Martinvl (talk) 20:42, 29 June 2009 (UTC)

I apologise for the delay in responding to this. I assume that you are referring to my deletion of the following text
IBAN wire transfers are not completely free of vulnerabilities. Every intermediate bank that handles a transaction can take a fee directly out of the wire payload (the assets being transferred) without the account holders knowledge or consent. There is no legislation or technical means to protect customers from this practice. If bank S is the sending bank (or brokerage), and bank R is the receiving bank (or brokerage), and bank I1, I2, and I3 are intermediary banks, the client may only have a contract with bank S and/or R, but banks I1, I2, and I3 can (and often do) take money from the wire without any direct arrangement with the client. Clients are sometimes taken by surprise when less money arrives at bank R. Contrast this with checks, the amount transferred is guaranteed in full, and fees (if there are any) can only be charged at endpoint banks.
I am not engaging in censorship and I rather resent the fact that you should think so. There is no such thing as an "IBAN wire transfer" because it is not a system for transmitting funds (i.e. it is not like Fedwire or ACH in the USA). You write this as though you believe that the IBAN's purpose is to protect the receiver of funds from meeting the costs of transmission. It is not. Its purpose is basically 1) to add a common system of check digiting to guard against data corruption and misrouting and speed up the process os sending funds by allowing errors to be detected early on and 2) to make it easier to identify all the information needed to route a payment by providing a common format for expressing an IBAN.
What you are referring to is of course an issue to receivers of funds transmitted across borders and between banks where there intermediate banks bearing costs of transmission. When payments are sent across national borders it is common for banks to be able to express who bears the cost of money transmission. This is expressed as a specific code in field 71a of the SWIFT transmission. Where this is expressed as /BEN it means that the beneficiary should bear all costs. If /OUR it means that the initial remitting bank will bear all costs (and will presumably pass this back to the sending customer). Nearly all banks use SWIFT and understand this. See as an example of this. Anyhows, the purpose of the IBAN is not to protect the parties from having to bear costs, but field 71a in the SWIFT message is how remitting banks convey instructions for this. Some words of warning tho! Inserting a value in field 71a adds greatly to the cost because it often prevents STP ("Straight thru processing"). When costs are directed to a particular party the costs are likely to be much higher than they would have been had the banks followed their normal practice. --Hauskalainen (talk) 18:43, 1 July 2009 (UTC)
FYI cross border payments were once very expensive within Europe. For instance it used to cost 24 euros to send just 100 euros from one account to another. The EU intervened and told the banks to reduce their costs. As a result the banks quicky set up a more efficient system and it now costs just two and half euros to make that same transaction today. In my opinion that is still too high, but its many times better than it once was!--Hauskalainen (talk) 18:53, 1 July 2009 (UTC)
The section on Security and Privacy looks put of place. It should either be removed or expanded. If it is expanded, I suggest that it be renamed "Commentary" and that matters such as security, privacy, reliability etc be discussed, a single sentence for each, using the pre-IBAN system as a point of reference. Any comments? Martinvl (talk) 07:58, 3 July 2009 (UTC)
User:Hauskalainen is having a lone battle against people using the Internet to validate IBANs. He might be right, on the other hand he might have an unreasonable paranoia about this. If there is a genuine problem with using internet-based IBAN validators, the responsible course of action for Hauskalainen is to cite references warning against this and to ensure that the message is distributed to all the different language versions of Wikipedia. If he cannot find any such references, then we must assume either that he has an unreasonable paranoia or that he is trying to publish original research � neither of which have a place in Wikipedia.
Prior to writing the material that Hauskalainen removed, I did a survey of the IBAN article in Wikipedia in a number of different languages. I tested all the IBAN validators that I came across using the fictitious IBAN example in the English language text. Am I not allowed to do that? If not, why not? Moreover, in my latest version, I stated Hauskalainen�s case and left him a space where he could insert his citation.
I invite other users to comment. Martinvl (talk) 20:50, 7 January 2010 (UTC)
I am not saying you are not allowed to do anything. Just that it is a wise precaution to be aware of the risks. I agree that I have no source saying that it is an unwise thing to do and I do find it amazing that even pament organizations such as APACS have put validators on their web site. Having said that I am sure that this is out of absentmindedness. Almost certainly the web site is managed by the IT department which no doubt had reason to develop the software and thought it a good feature for the web site. Personally I think the more security conscious,if they thought about it, would take a different opinion of the wisdom of encouraging people to "validate" information by giving confidential information to third parties who do not need it and at worst could misuse it. You would not do this with your credit card details so why do it with your bank account details? I have a professional degree in banking and have been responsible in banking organizations for both payment systems and systems security. I do not make this case on the basis of whim, but in the spirit of encouraging people to treat their personal information with some degree of care that does not expose them or their banks to unnecessary risk. I wonder why you would want to remove such a warning. It does no harm to think about the risks.--Hauskalainen (talk) 14:41, 23 January 2010 (UTC)
If you are in the banking industry, then you have a better platform than most to let your fears be known. Your energy is best expended there rather than in Wikipedia. Once you have convinced the banking indutry of your fears, by all means reference the public advice given by the banks in Wikipedia. Until then please abide by WIkiepdia's policies. Martinvl (talk) 15:37, 23 January 2010 (UTC)

Proliferation of links

I have become a little concerned at the proliferation of links and references that are appearing in this article. I propose pruning then down as follows:

  • External links will be limited to official definitions only (SWIFT, UN agencies etc).
  • References to documents issued by individual country�s national banks will be removed unless that country is not yet in the official SWIFT list.
  • References to parts of an IBAN will be replaced by Wiki-links to the language concerned � for example the reference of a CIN (for Italy) will be replaced by a Wiki-link to it:Coordinate bancarie, but will appear as CIN in the IBAN article. A warning will be posted that some of the Wiki-links will be to non-English articles.
  • References or links to any particular bank will be regarded as SPAM unless it is used to illustrate or to back up a specific point that is described in the article�s text which is not otherwise illustrated or backed up.

Any Comments? Martinvl (talk) 11:27, 18 January 2010 (UTC)

No original though or research

Wikipedia has a policy of no original thought or research - WP:NOP.Hauskalainen, but adding reams of unreferenced warnings about the dangers of publishing one's IBAN appears to be promoting is own views. The following is taken from the Wikipedia policy on the matter Wikipedia does not publish original research or original thought. This includes unpublished facts, arguments, speculation, and ideas; and any unpublished analysis or synthesis of published material that serves to advance a position. This means that Wikipedia is not the place to publish your own opinions, experiences, arguments, or conclusions.

When adding his unsourced material, User:Hauskalainen removed a citation needed flag, but failed to add a citation. Another of Wikipedia's policies is that of verifyability WP:VERIFY. Removal of citation needed flags should only be done once a reliable citation has been added. (See Template:Citation needed) Martinvl (talk) 10:12, 23 January 2010 (UTC)

For reference, I am close to but not part of ECBS: I smoothed their way, for instance, in getting Malta (the first small country, non-EU at the time) on board, identifying systemic failings which they have rectified. —Preceding unsigned comment added by (talk) 10:25, 17 August 2010 (UTC)

User:Martinvl is a vandalist to me

See my version

-- (talk) 10:40, 23 April 2010 (UTC)

-- (talk) 10:52, 23 April 2010 (UTC)

I don't understand why I have got so many clones and tails -- (talk) 10:54, 23 April 2010 (UTC)

At a high level your additions looked like advertsing to me. At my place of work, I was denied access to yoru website as being "inappropriate". Also, please identify yourself. Martinvl (talk) 11:58, 23 April 2010 (UTC)

See what this person did....too

User:SmackBot —Preceding unsigned comment added by (talk) 10:48, 23 April 2010 (UTC)

Clarification Needed

I removed this flag as I cannot see why it was put there. If the original editor (who is hiding behind an IP address) wishes to reply, please explain what sort of clarifiaction is needed. Martinvl (talk) 11:58, 23 April 2010 (UTC)

Subsection: Calculating the modulus of a large number

I came here to find out how to verify an IBAN (i.e., the modulo 97 = 1 algorithm). As explained in the named subsection, this calculation cannot be performed easily on a PC as integers with so many digits are not handled correctly. The named subsection aims at providing a solution to this problem. Unfortunately, I fail to properly apply this information. I think it would be good not only to provide the formulae that form the basis of the smarter approach but also to give an example based on a sound IBAN. Tomeasy T C 22:02, 28 July 2010 (UTC)

Did you download the program for which a link is supplied? Martinvl (talk) 11:36, 29 July 2010 (UTC)
I do not think a program would answer my question, neither solve the issue that I am pointing at, which is that we provide a lot of maths in one subsection which is too difficult to apply without an example.
Moreover, which link do you mean. there is no download offered in the section I am talking about. Tomeasy T C 19:24, 29 July 2010 (UTC)
There are links in the subsection "Online IBAN validation". Martinvl (talk) 21:11, 29 July 2010 (UTC)
Are you willing to improve the article?
I know there are plenty of websites where one can check an IBAN. This is not why I started this thread. I want this article to explain which criteria are checked, and I think the article does a good job in doing so in the subsection International_Bank_Account_Number#Calculating_and_validating_IBAN_checksums.
However, the checksum computation, as described in that subsection, cannot be conducted with ordinary PC applications. To this end, modular arithmetic is used to convert the check into a form that can be handled more easily. The article describes this procedure in the following subsection International_Bank_Account_Number#Calculating_the_modulus_of_a_large_number. I endorse also this subsection, but I think it requires improvement as I failed to put it into practice. It is to abstract and lacks any obvious connection to the IBAN problem that the article is supposed to treat.
Now, do you understand the subsection containing modular arithmetic, and can you help to improve it, can you enable readers to conduct the checksum themselves on their PC? If so, let's get it done. If no, please stop doing as if things were perfectly fine by actually not addressing the issue with short references to unrelated content in other sections. Tomeasy T C 22:53, 29 July 2010 (UTC)
Hi User:Tomeasy
I was disappointed to see you revoking my changes, especially when they were preliminary to additions that you had asked for. Firstly, I felt that the section detailing how to calculate the modulus of a large number was a little out of place in the IBAN article, so adding a tutorial section would have made it even more out of place. For that reason I felt it best to have that sort of detail at the end of the article. May I draw your attention to the fact that the "Notes" section was specific to the table that preceded it, not to the article as a whole. Martinvl (talk) 18:30, 30 July 2010 (UTC)
Fine. I am not against moving the section down and I understand the confusion about the notes. I reverted for a second reason: You deleted 480 bytes. What did you delete and why?
More important that the location of the section, to me, is the content. Can you help this? Do you understand it so much that you can give some more explanation. 18:37, 30 July 2010 (UTC)
When I originally rewrote sections of the article, I converted the old text into comments so that I could quickly reinstate them if what I was writing was incorrect. Nobody has objected, so when I moved the text, I deleted these comments as they no longer served any purpose. I think that this explains what I am doing, so I will reinstate my changes. Martinvl (talk) 20:17, 30 July 2010 (UTC)
Fine again. Now, can we do something about the issue itself, or are you as inapt as I a to make sense out of this subsection? Tomeasy T C 21:53, 30 July 2010 (UTC)
Done - BTW, I picked up a typo in the maths while I was doing it. Martinvl (talk) 07:20, 31 July 2010 (UTC)

Excellent! Just a pity that so many comments were posted without you showing that you are able and want to provide this excellent work.
I think I found a mistake, but you might be more apt to answer. Instead of "Column di are the progressive values that are calculated using the formula di=mod(di-1,10)." Should it not read "d_i=mod(d_i-1 x 10, 97)"? At least, this is what I observe in the respective column.
Also, instead of "Total" I would like to write "Sum (a_i x d_i)" if this is indeed what is done there. Unfortunately, total is not fully explanatory.
Is there any specific value in the sentence "In practice, these numbers are hard-coded into the relevant computer code."? Did I miss something here that makes the treatment of this item intrinsically different from the other items. with my current understanding, I would delete the phrase is not needed to explain how this cell value is computed. Tomeasy T C 10:32, 31 July 2010 (UTC)

Hi User:Tomeasy. I think that you are right on both counts. I have made the changes and removed your comments. The reason that I added the phrase about hard-coding of the numbers is that all the programs that I have seen have these numbers hard-coded as an array - something that I would do as well if I was writing a performance-critical program (in real life. I have however modified the sentence. BTW I am a IT Engineer with many years of experience of writing computer programs). Martinvl (talk) 11:55, 31 July 2010 (UTC)
I implemented the change to "d_i=mod(d_i-1x10, 97)" and some more copy edits. I also tried the algorithm myself on Excel with some IBANs, and it works for me. So, I am actually done here. Perhaps it would be a good idea to add a reference to the modular arithmetic modifications in the previous subsection, to prevent all this looking like original research.
Thanks a lot, and happy editing. Tomeasy T C 12:48, 31 July 2010 (UTC)
Hi Tomeasy - You are the first person to look at the code critically, even though the article gets over 1000 hits a day (especially Mon-Fri which suggests that most hits are business-oriented). I will certainly add a few more references and I will alos look out for a picture or two. Regards. Martinvl (talk) 13:20, 31 July 2010 (UTC)
Now that I understand at least how to execute the formula generated by modular arithmetic, I can also ask something about the theoretical subsection:
Instead of should it not read ? Smilarly, i cannot make much sense out of the formula that follows right afterward. Or perhaps, do a_i and d_i have different meanings in the two subsections?
Also, in the theory section I would starting counting with 1, so it is in line with the example. Tomeasy T C 15:32, 31 July 2010 (UTC)
User:TomeasyI think it important that there is consistency throughout the article and I welcome the attention that you have been giving. I have made a few changes to keep a consistent notation for the mod operator - previously I had used both mod(A,B) and A mod B. I have now standardised on A mod B. Martinvl (talk) 15:58, 1 August 2010 (UTC)
Thanks for the words. Your last edit was certainly an improvement. Things that catch my eye at the moment are:
  1. The subsections on validating and calculating are highly repetitive.
  2. There's a section IBAN validation and a subsections on Validating IBAN checksumswhich are highly repetitive and should be merged.
  3. The use of the word checksum. Do we mean the whole validation process including length as per country or just the modulo calculation?
  4. Why do we write mod bold everywhere?
  5. The repeated not The proof of this is an undergraduate maths level problem does nothing else than intimidate and talk down to people. If the proof is so simple rather use a note to show it.
  6. ECBS needs expansion.
  7. Why do we sometimes write Mod-97-10 and not simply Mod-97?
  8. Personally, I am not a big fan of using x as the sign for multiplication and we are also not doing it consistently. I prefer no sign between symbols and centered dots between numbers.
  9. Shouldn't we exchange d_i and a_i, because d_i sounds more like digits of D than a_i?
I will make some more changes. Feel free to take them back or further improve where it appears appropriate to you. Please let me know what you think about the points. Tomeasy T C 18:31, 1 August 2010 (UTC)
Hi Tomeasy. Thank you for your comments. I have a few responses
  1. Sorted by you.
  2. I have changed the wording slightly. I would prefer to keep them apart - the frist is aimed at the general public and is written in language that the layman can understand - the second is aimed at geeks (like you and me).
  3. I have modified the wording. I hope that it is clearer now.
  4. I am happy to write mod instead of mod. I am not happy with "mod" when using ordinary text.
  5. This is an uneasy compromise. The proof is not trivial, I do not have a source for it and no mathematician would class it as WP:OR, though some Wikipeduia editors who have very little background might do so. Writing out the proof would not add anything to the article.
  6. Sorted by you.
  7. Mod-97-10 means "Mod 97 using numbers to base 10". It could be reduced to "Mod 97" if you think it an eyesore.
  8. I am quite happy with your suggetion, apart from when we are multiplying numbers.
  9. Maybe it would be better to inerchange A and D (it is easier!).
I have two other items that I would like you to comment on
10.I think that the map would be better placed in the section "Geographical usage" rather than "List of valid IBANs by country". Any views?
11.I am thinking of generating a typical header of a British bank statement (based on a real bank, but all the data annonymised and the bank being called "The Wessex Bank" (so that it can use the routing code WEST which is used in the examples in the article. Any comments? Martinvl (talk)
2. Yes, you improved and now as I understand the intention (and approve it), I will perhaps copy edit it accordingly. I have not spend a lot of time yet on the layman's section, perhaps I can make it even easier.
3. I see that the source document actually uses check digits. Perhaps we should do the same. This is also clearer in the sense that it relates only to the Mod-97 calculation on the two digits.
4. I think I will make this change. Bold gives undue weight, but yes it should not be regular text as it is not regular prose.
5. If it is not trivial, this condescending statement is even more wrong. I would like to delete it then. Perhaps we write in the text that "it can be proven that".
7. I see that the source does so and I thought it relates to the decimal system. I think we should not use this level of detail, it is more confusing than anything. I am not afraid that anyone would come to the idea that this section is talking about binary or hexadecimal calculations.
8. So you would not like to see 27·10? Would you stick with 27x10 or do you have another idea. I will implement the changes for the multiplication of symbols.
9. That would not really help as digit starts with d and not with a. By using the replace function of the editor, it is actually no problem to change many instances at one ;-)
10. Absolutely, that's where it ought to be.
11. Sure, that would be a nice addition. It would function as an anchor point for the readers to identify with the topic of this article. Tomeasy T C 22:47, 2 August 2010 (UTC)
Hi, I have implemented a few of the changes suggested above. I notriced that the ISO 13616 stanard explicvitly stes "MOD 97-10", so I have included that but with a qualifier. I am still thinking about points 8 and 9. Are you happy with the change of emphasis regarding Note 5.Martinvl (talk) 12:02, 3 August 2010 (UTC)
Yes, I think that is a good solution. We mention MOD-97-10 once upfront like it is mentioned in the standard. Henceforth, we use the simpler formulation.
5. To be honest, I do not really understand the purpose of stating that the proof would be standard classroom. Which standard? and even if that was clearly defined, what does the reader learn? Does it make the claim any more reliable? Is the purpose to stop people asking and think it will be alright? I think it just sends people who did not like math classes. they will come to the conclusion that this thing is beyond them. And those with some understanding will also get a bad feeling as they might not be able to do it just like that, and then we tell them that it is standard. Why do we need to say anything? We have a reliable reference that makes this claim. So, we report it.
8. We could just remove the "x" for the symbol multiplications, but leave them when multiplying numbers.
9. Why do you see a problem in changing a and d. Is there a common standard for the way we currently chose them? Do all the textbooks use them like this? My intention for the is simply that a formula is easier to understand if you have to spend less energy on remembering which symbol meant what. For this purpose, I find d fr the digits just useful.
I ruled out all list items that we are done with, all other ones still require some attention. Tomeasy T C 17:15, 3 August 2010 (UTC)
Hi Tomeasy. I have sorted out (5). When I said that I was "thinking about (8) and (9), I wanted to print the article out and look at it on the train on the way home. I hope to do that tomorrow!. Martinvl (talk) 21:07, 3 August 2010 (UTC)
I implemented the change proposed in (3), and did some copy editing. Take back the changes you do not like. If I disagree somewhere, I will comment here. Tomeasy T C 22:20, 3 August 2010 (UTC)
Are you still thinking about (8) and (9)? Otherwise, I would implement the changes I proposed.
Hi Tomeasy. I am quite happy for you to go ahead with (8) and (9) - I am afraid that real life got in the way. Regards Martinvl (talk) 18:57, 5 August 2010 (UTC)
Ok, they are off the list now. Actually, I did not remove of the X multiplication signs. It did not really work out well. Instead, I added one where it was missing for consistency. I think we can take a break now :-) Tomeasy T C 19:53, 5 August 2010 (UTC)
I agree that a break would be a good idea, apart from the specimen bank statement that I am creating. It will be annonymised (for obvious reasons), but will be based on the designs used by both the Halifax and Lloyds Bank. I will call the bank "Wessex Bank plc" and the customer will be "Mr Arthur King, 1 Camelot Drive, Avalon, Wessex". I will find an apprpriate fictitious postcode. The bank account and details will be those used in the examples (whic is why I tried to incorporate "WEST" into the bank name somehow. Martinvl (talk) 20:21, 5 August 2010 (UTC)
Good luck with it! And thanks again for helping me out with the modular arithmetic algorithm. Tomeasy T C 20:35, 5 August 2010 (UTC)


BIC and IBAN are used by banks in Madagascar. I have used them myself sending money there this year. I don't know why Madagascar does not appear in the IBAN registry. Hawklord (talk) 20:18, 9 August 2010 (UTC)

Hi Hawklord,
I spotted your querry a few minutes ago. There might be a very simple explanation - please publish the IBAN that you are using, keeping the first four characters as they are, but to preserve the privacy of your recipient, please replace the subsequent alpha characters with a "Z" and subsequent numeric characters with a "9". Thus the example IBAN used in this article would be written: "GB82 ZZZZ 9999 9999 9999 99". Martinvl (talk) 10:13, 20 December 2010 (UTC)

Reverted IP edit

Hi Martinvl, You reverted an IP edit which I think was pretty smart and certainly not incorrect as you said. The way they put it makes the check digit calculation easily possible with normal calculator in only 5 steps. What we are currently showing in the example section is easily done if you have a spreadsheet calculation, but it comes in less handy than the IP proposal. Tomeasy T C 15:31, 17 August 2010 (UTC)

Hi Tomeasy,
The edit that I deleted was erroneous in that the author forgot about the column ai in the step-by-steop example in the article (at least I could not find anything that looked like ai). Martinvl (talk) 19:30, 17 August 2010 (UTC)
Bottom line for me is that the algorithm described works, and I think it was actually the intention not to express the algoritm in terms of a_i etc.
When you say it was erroneous, do you mean it does not work? Or do you mean it is not equivalet with the algorithm shown? Tomeasy T C 20:14, 17 August 2010 (UTC)
WHat he wrote would not have worked. What he should have written (my additions are in bold) was In plain English, therefore, calculate the Mod 97 remainder of whatever length at the start of the string you can handle. Multiply it by an appropriate weighting factor and Tack it on to the start of the remainder and repeat until you have finished. Then calculate the Mod 97 remainder of the number that you have. Martinvl (talk) 20:56, 17 August 2010 (UTC)
I tried the algorithm as described by the IP, without your additions, and it works fine - at least for the example case in the article. The odds that it works just coincidentially are (maybe?) 1:97.
I think it deserves some thinking from our side. Or even better the IP joins in the discussion. Perhaps the proposal is not equivalent with our algorithm on a first glance, it it works without the static array, and does not go digit by digit but as many digits as wanted at each step. If this was a working algorithm, and can be proven as such, it would be favorable in some aspects. Tomeasy T C 22:52, 17 August 2010 (UTC)
I am unhappy with the term "tack it on". What exactly does IP mean? Martinvl (talk) 06:06, 18 August 2010 (UTC)
I am certainly not happy with the language the IP used, and I would give it a thorough copy edit if necessary. But right now this is not the point. Is there a factual problem with the addition? You called it incorrect at first, so I guess you have something to say. Perhaps you can answer my questions above. This might bring us a step further. Did you try it out yourself once? And why did you remove the title of this section header? Tomeasy T C 17:28, 18 August 2010 (UTC)
Hi Tomeasy, Firstly, my removal of the section title was an error on my part. I was doing a copy and paste for teh edit summary and did a cut and paste.
Regarding checking out IP's example - I checked the maths behind his method and that proved him to be incorrect - unless I missed something which si why I would like to see a worked example so that I know what he really meant. Martinvl (talk) 17:34, 18 August 2010 (UTC)
Try it with the IBAn we have in the article. For me it works, and I doubt this is coincidence. Tomeasy T C 20:09, 18 August 2010 (UTC)
Hi Tomeasy. Before I spend any time trying out IP's algorithm. I want to know what he meant by "tack it on". I would also like to see your worked example to satisfy myself that it is mathematically sound. BTW, I am expanding the worked example. Martinvl (talk) 06:32, 19 August 2010 (UTC)
Sorry Martin, I did not know that you do not know how to apply it. I'll try to explain:
Initial question: 3214282 91234 56987654 32161182 mod 97 = 1? The IP suggests that this is equivalent to doing the following:
321428 mod 97 = 67, instead of length 6, you could have chosen any length that you wanted/ that you can handle.
Prepend the 67 to the remaining string and you have an equivalent formulation of your initial question: 67912345698765432161182 mod 97 = 1. This may stil be too large a nuber so you can repeat the step.
6791234 mod 97 = 70 leads to 7056987654, then again:
7056987654 mod 97 = 4 leads to 432161182, and finally:
432161182 mod 97 = 89.
OMG, I does not work! Shame on me. Let's see if I can find again the sheet of paper where I did it last time (with smaller steps) when the result came out fine. Since the length is allegedly indifferent, there are many possible ways of chopping down the large number. Nevertheless, this would only show that by then I made a mistake or just tried a lucky combination. The algorithm as proposed, however, is not an equivalence formulation of the problem. Tomeasy T C 07:29, 19 August 2010 (UTC)
Thanks God, I found my sheet back. Well, the algorithm works (coincidentially) if you chop the large number in numbers with lengths: 3, 5, 8, 6, 7, 7. Note that the sum of these numbers is 36, i.e., the length of the large number. So, I hope it can be understood what I did. The finaly calculation was 2461182 mod 97 = 1, q.e.d. But this just shows that the algorthim worked for one special case, hence, no proof for the actual algorithm which is meant to work for all IBANs and all length sequences. This, on the contrary, has been proven wrong by virtue of an example, given above, where the algorthim failed. Tomeasy T C 08:40, 19 August 2010 (UTC)

Do you think there can be a way to solve the question is a reduced number of steps as proposed by the IP, with the only difference that it works?Tomeasy T C 07:31, 19 August 2010 (UTC)

Hi Tomeasy, I don't know if you have done any computer programming - I have been programming computers for a living for nearly forty years. Although the method that we used looks as though it has a large number of steps, in reality it contains very few steps, but these steps are repeated a number of times within a "loop". I don't know if my second column helped. If not I will happily remove it. Also, I will think about adding a short preamble to describe in words what is being done. Martinvl (talk) 14:18, 19 August 2010 (UTC)
You showed this attitude already in the previous thread. Please be assured, whether it matters to our discussions or not (I'd say it doesn't), that I have a lot of experience in programming.
Your thinking from this angle brings you to wrong conclusions. I was never thinking that the algorithm shown in the article poses the slightest problem to a programmer, and your allegation that I would think so upsets me. I guess, you would not come to think this way if you would just assume that other people might know what they are talking about as well.
Instead, you might have come to think of other advantageous of the (not working) algorthim that I had in mind. In deed, a coded program would not worry about 40 or 4 simple loops (if called once). However, in my previous edits, I was talking about the advantage of doing the check yourself, with a calculator only. The IP algorithm is easily done in 2 minutes, while the article's algorthim is very tiring if you do it yourself. You would need to concentrate immensely not to make a mistake and it would take at least 5 minutes. This is why the IP's algorithm, had it worked, appealed to me.
I am sorry that this probably sounded curt to you, but I feel offended by your constant repetition of your experience coupled with the insinuation that this is not mirrored on my side. Tomeasy T C 14:46, 19 August 2010 (UTC)
Hi Tomeasy, I can assure that that no slight towards you was meant whatsoever. I was at fault for not having read your user page more carefully and not seeing that you are a C++ programmer for which I apologise. (Maybe I should place a few more notes on my own user page).
I think that it woudl be relatively easy to write an algorithm that one can do in a few steps on a calculator (or spreadsheet). The algorithm should take no more than six lines. I will come back in the next day or two as I real life is calling. Martinvl (talk) 15:43, 19 August 2010 (UTC)

Here is the algorithm:

Assume that we want to find
eeeeeeeeddddddddccccccccbbbbbbbbaaaaaaaa mod&nbsp97
where eeeeeeee is an 8 digit sequence, as is dddddddd etc.
The solution is
E = (eeeeeeee mod 97) x 61
D = (dddddddd mod 97) x 75
C = (cccccccc mod 97) x 62
B = (bbbbbbbb mod 97) x 81
A = aaaaaaaa mod 97
Answer = (A + B + C + D +E) mod 97.
The factors 81, 62, 75 and 61 are mod(10^(8n),97) where n takes on the values 1, 2,3, 4

Is this simple to follow? BTW, I checked it on a spreadsheet. This is almost identical to the algorithm described, but it uses eight digits at a time, not one. Martinvl (talk) 18:04, 19 August 2010 (UTC)

Thanks for your efforts and coming up with a workig algorithm that addresses the issue. I think we are fine with reporting on the article just the algorithm that we currently have. Tomeasy T C 16:19, 20 August 2010 (UTC)

IBAN Example

An annonymous editor has added axamples for all the IBAN entries in the list of entries. While such additions might be interesting, they do compress the other columns to the extent that I think this negates their value. Since they do not have the correct format (a space after every fourth character) and adding this format will render the table even less readable, I must question their value. Moreover earlier sections explain the IBAN structure so an example for every single country adds very little, if any, additional value. Martinvl (talk) 21:33, 26 April 2011 (UTC)

There are examples in the Background section already. I would argue that this is sufficient in order to illustrate the concept. Tomeasy T C 23:26, 26 April 2011 (UTC)

under 02

(Moved to end of talk page by User:Martinvl) if i create the iban check digits with modulo 97 and get 98, the check digits would be 00. why are the check digits named by the bank 02. i knew that 00 is incorrect for every iban. but what is if i create digits that would be under 02 ??? the iban named by the bank is DE 02 600501017495530102 (talk) 09:03, 8 July 2011 (UTC)

The article states "Theoretically, it is equally possible to generate check digits in the ranges 00 to 96, 01 to 97, and 03 to 99, but the standard is silent as to whether or not these ranges may be used.". Thus both 02 and 99 are valid check digits for the same account number. The IBAN specification does not specify which is correct - the bank can choose. BTW, 98 mod 97 is 01, not 00. Martinvl (talk) 11:24, 8 July 2011 (UTC)

Reliabiliity of non-IBAN routing

I removed the newly added citation Citation? Wiring funds to American bank accounts is done with the SWIFT and ABA routing number + account number. If the name on account mismatches, the wire bounces back. So the reliability is similiar. on grounds that this is not true. If the user mistypes the IBAN, the software in the user's own computer will trap an invalid IBAN and request that the user enter a correct IBAN. The message with an invalid IBAN will never get to the wire which is why the reliability is enhanced. Martinvl (talk) 06:00, 9 July 2011 (UTC)

Can kk ever be a "national check digit" as the table seems to indicate?

The larger table containing formats for each nation seems to imply that the check digit can be a national check digit. But how can this be? The check digit is arrived at by taking the national details, wrapping it up into the IBAN format and THEN calculating the check digit pair. Some national account number systems included a check digit before IBAN was introduced but this national check digit is within the BBAN and is not the additional check digit pair designated as kk. If I am right could someone be kind enough amend the table to remove kk being designated as the national digit. (talk) 00:26, 3 March 2012 (UTC)

In the table, lower case "kk" is used to identify the IBAN check number and upper case "K" or "KK" to identify the national checksum where applicable. Martinvl (talk) 07:27, 3 March 2012 (UTC)

Character Set

I have reinstated the note about the character set - the example chosen for the table specifically included ONE country that uses the Greek character set, ONE that used Hebrew, ONE that uses Arabic etc. Every now and again somebody adds their own coutnry, but to keep things under control, I remove it. I justify the removal by referring to the end-note — Preceding unsigned comment added by Martinvl (talkcontribs) 10:30, 27 March 2012‎

We shouldn't allow article content to be dictated by editing conflicts. If the characters used need to be specified then so be it, but there is no need to patronise the reader by pointing out individual instances of countries which adhere to the standard despite not using the character sets in question in their own writing. Chris Cunningham (user:thumperward) (talk) 10:46, 27 March 2012 (UTC)
On the other hand it is not a bad thing to remind those readers who only use the latin alphabet that other alphabets exist. In the past I have removed the Turkish example (I think they objected to Greece beign there). Martinvl (talk) 11:50, 27 March 2012 (UTC)


I was never happy with the detail to which the algorithms were generated, but I cooperated with another editor in writing the example. Following on Thumperwald's suggestion, my feeling is that the first two subsection on Validation should be retained and the last two moved to Wikiversity with appropriate links between the two. Any comments? Martinvl (talk) 10:43, 27 March 2012 (UTC)

Sounds like a plan. The content is good, but ideally this article should concentrate on why the standard was created, what problems it solves and how it has been adopted rather than technical implementation details. Chris Cunningham (user:thumperward) (talk) 11:00, 27 March 2012 (UTC)
I will add this to my "to do" list. Martinvl (talk) 11:51, 27 March 2012 (UTC)

Changes made by Canepa

Canepa's request for a citation was made in respons eto a comment that I made at Wikipedia talk:No original research. Since I volunteered this information, I regard Canepa's rejection of my assertion as being bad faith, especially as this is the first edit that he made to Wikipedia as a named editor. Please discuss further at Wikipedia talk:No original research. Martinvl (talk) 12:44, 6 August 2012 (UTC)

I was following-up the admission you made, in WT:NOR, that you attempted to excuse the absence of a reliable source for a stated fact, yes. Please give your reasoning for assuming that WP:VER does not apply to that contribution. Canepa (talk) 13:03, 6 August 2012 (UTC)
Please discuss this at Wikipedia talk:No original research, not here. Martinvl (talk) 14:43, 6 August 2012 (UTC)

Actually since you are arguing about a concrete text it should be discussed here rather than there, but you can post a a notification of this discussion there.--Kmhkmh (talk) 15:21, 6 August 2012 (UTC)

Having looked that text in question now, I don't a specific citation is necessary here. The knowledge needed to fully understand the description can be found in most number theory primers (modulo computation), the only other things that is used are the positional properties of the decimal system. One could could cite a number theory book for the modulo computation or may some textbook that covers the IBAN, but I'd agree with Martinv1 everything here is basic enough that a citation is not really required here. Moreover modular arithmetic is linked, so that readers can look up the basic properties there (and sources for it).--Kmhkmh (talk) 15:43, 6 August 2012 (UTC)

I have reworded things slightly so that the computer program is itself the reference and the text explains the reference. Martinvl (talk) 16:29, 6 August 2012 (UTC)
But you didn't supply the required reference, so I've put the request tag back in there. There are 2 options: supply a ref to a reliable source or remove the content. For further discussion on this see WT:NOR#Original mathematical proofs are not interpretations. Canepa (talk) 08:39, 7 August 2012 (UTC)

Under referenced and too much reliance on primary sources

I've tagged the article as requiring more references and as relying too heavily on primary sources because, of the few references provided, the majority are primary sources. Canepa (talk) 09:11, 7 August 2012 (UTC)

I have removed the "RefImprove" flag. Would the editor who placed that flag please identify the sections that need improving. If such sections are only in a small section of the, please highlight the specific sections - the banner across the top of the page being totally unneccessary. Martinvl (talk) 09:30, 7 August 2012 (UTC)
None of the sections are adequately referenced. The banner is appropriate. Canepa (talk) 09:40, 7 August 2012 (UTC)
Section by section, the main problems are:
  • Background: para 1 is unreferenced, para 2 is partially referenced and only to primary sources, para 3 is partially referenced and only to primary sources.
  • Practicalities: para 1 is unreferenced, para 2 is partially referenced and only to primary sources.
  • Features: para 1 is unreferenced, para 2 is partially referenced and only to primary sources, para 3 is unreferenced.
  • Algorithms: Largely unreferenced.
  • Adoption: Largely unreferenced.
Canepa (talk) 09:53, 7 August 2012 (UTC)
There is nothing wrong with using primary sources per se - if you read WP:OR, it is the way in which they are used. Since this is an article about the use of a standard, then the document that describes the standcsard is the authoritative source. Martinvl (talk) 10:18, 7 August 2012 (UTC)
I have removed the banners again. This is an old article, and particularly the background section was written before Wikipedia had strict rules about in-line citations and long before I became associated with Wikipedia. It is extremely difficult to find the original text - when I did a Google search, all that I got were hundreds of banks who reproduced this paragraph verbatim. I don't know if they got if tom Wikipedia or if they got it from the same source that Wikipedia used. You would be a lot more useful if you tried to find where this text came from. Remember that on weekdays this article is accessed over 3000 time a day and you are the first person to demand a source. Either you know something that the rest of us don't (if so, please share it with us), or you are just being anal. Martinvl (talk) 10:58, 7 August 2012 (UTC)
Read each sentence in the article, one at a time, and for each, ask yourself whether what you have just read can be readily verified by the reader using a referenced source. For all the sentences in the article where the answer is "no", a reference is still required. I did that, and I believe that there are sentences in each section which require referencing. Primary sources are fine for pure objective facts, but for anything subjective, possibly including the reason why an objective fact is relevent, a reliable secondary source should be provided. I don't remember the Wiki policies exempting older articles. If sources cannot be found, the content must be deleted. Canepa (talk) 11:06, 7 August 2012 (UTC)
Please read Wikipedia:Citation overkill and Wikipedia:You don't need to cite that the sky is blue. Also, I will go through your objections shortly, but I do have a real life. Martinvl (talk) 11:16, 7 August 2012 (UTC)
Neither of the two essays you cite are relevant in this case, the policy WP:V is what is important here. I am happy to leave the banners in place for a few days before deleting any still unreferenced content, but I not happy to leave the article as it is, with a large proportion of unreferenced content and with no banners to warn readers. If you demonstrate good faith by returning the banners, I will take no further action for a few days, to allow some time for sources are searched for. Canepa (talk) 11:31, 7 August 2012 (UTC)
Now it's about time for a full stop, if you start deleting content here, despite what was explained to you by several other people here and at Wikipedia_talk:No_original_research then I'd consider that as vandalism. Yes WP:V applies here, but as it was explained to you you are reading WP:V wrongly in the context of the problem at hand. If you don't trust the judgement of Martinv1 or me request an opnion from another editor with domain knowledge at Wikipedia talk:WikiProject Mathematics or a similar project site (computer science, physics, third opinion).--Kmhkmh (talk) 12:14, 7 August 2012 (UTC)
We only discussed one of the many, many, missing references at WT:NOR, and it was one of the ones relating to the math. Most of the several other unreferenced statements are not mathematical. Read the article and the discussion here, then come back if you believe I am wrong in expecting this article to be made verifiable. Canepa (talk) 12:28, 7 August 2012 (UTC)
Well yes that comment only referred to modulare arithmetic issue as far as the overall sourcing is concerned that could be improved indeed. Still it doesn't look like an article being particular problematic, personally I prefer to keep (top level) tagging to articles where there is a serious issue.--Kmhkmh (talk) 13:53, 7 August 2012 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── Here is my answer to Canepa's comments. For the sake of clarity I have reproduced his comments and italicised them:

  • Background: para 1 is unreferenced, para 2 is partially referenced and only to primary sources, para 3 is partially referenced and only to primary sources.
...Para 1 & 3 have been in the article largely unchanged for years (19 August 2005) and well before I became associated with Wikipedia. In trying to find a reference using Google, I came across literally hundreds of references using this exact text. Either they used the same reference as the original editor, or they used Wikipedia. You would be much more useful trying to find the original reference rather than pontificating about the lack of one. Meanwhile, if hundreds of banks are happy with that text, there is no need to flag it, obviously the customers are happy.
...Para 2 - These primary sources have been used in a perfectly valid way - to quote WP:PRIMARY "...descriptive statements of facts that can be verified by any educated person with access to the source but without further, specialized knowledge".
  • Practicalities: para 1 is unreferenced, para 2 is partially referenced and only to primary sources.
...Para 1 - I have added a citation to keep you happy.
...Para 2 - I do not understand your comment.
  • Features: para 1 is unreferenced, para 2 is partially referenced and only to primary sources, para 3 is unreferenced.
...Para 1 - If you read the rest of the article, you will see that this paragraph brings everything together. This is a case of WP:BLUE.
...Para 2 - The text is virtually a verbatim quote from the standard - See WP:PRIMARY
...Para 3 - Possibly a citation regarding the first sentence.
  • Algorithms: Largely unreferenced.
...See Kpmkpm's comments
  • Adoption: Largely unreferenced.
...I have added a reference to keep you sweet. The rest is legacy stuff from way back - probably written by a banker.

In short, therefore, apart from the algorithms section, I have answered all your questions. Martinvl (talk) 14:08, 7 August 2012 (UTC)

Just a general before we get a somewhat pointless primary/secondary source discussion. As far as as the technical part and (technical) application of banks are concerned primary sources would be sufficient, as secondary literature would simply reproduce that anyway and just provide additional context or nicer description.--Kmhkmh (talk) 14:21, 7 August 2012 (UTC)
It might also be worthwhile reflecting on the following from Wikipedia:What SYNTH is not - SYNTH is when two or more reliably-sourced statements are combined to produce a new thesis that isn't verifiable from the sources. If you're just explaining the same material in a different way, there's no new thesis. - in other words if all that I am doing in the algorithm part is explaining things, then I am not introducing any new ideas, so no citations are neccessary. Martinvl (talk) 14:31, 7 August 2012 (UTC)
Well no, you still need the original citations mentioned in WP:SYNTH. Yes you are not creating anything new and your simply explaining old material (hence not violating WP:SYNT_H), nevertheless you still need to provide references to that old material, enabling others to check your explanation without having to research all needed sources themselves.--Kmhkmh (talk) 14:42, 7 August 2012 (UTC)
In the sentence leading up to the algorithm, I have the text "... so many programs, such as the Javascript program published by ...". Would it be appropriate for me to rearange the text so that it is an explanation of the algorithms used by this program? It woudl be wriotten in such a way that the reader does not need to consult the program to understand the text. Martinvl (talk) 15:05, 7 August 2012 (UTC)
Using a program as source doesn't strike me as appropriate, instead use the document which were used to write the programs. "Reverse engineering" is not a good way to write and source encyclopedic articles. Also external links within the text part of WP article are usually frowned upon. You could put it in a footnote, but as i said before as source/explanation I wouldn't use it at all.--Kmhkmh (talk) 16:33, 7 August 2012 (UTC)
"In short" Martinvl, there are still much unreferenced content in the article. You may have added references, but you haven't addressed the underlying problem. If references can't be found for old content, then it must be removed. Canepa (talk) 07:58, 8 August 2012 (UTC)
If Canepa hasd the courtesy to read the rest of the paragraph before splattering it with "Citation needed flags", he would have seen that ISO 13616-1:2007 and ISO 13616-2:2007 are already referenced. If he had one iota of intelligence, he would have realised that this implies that the standard has been split into two sections. If he knew anything about standards, he would have realised that the history of the standard is documented in the standard. Martinvl (talk) 08:05, 8 August 2012 (UTC)
Look again, it wasn't the lack of a cite for the standards that I flagged. What I flagged was the need for sources (secondary will be required for the subjective content) to support:
  • "To overcome these difficulties, the European Committee for Banking Standards (ECBS) developed the IBAN"
  • "which was later adopted as an international standard under ISO 13616:1997."
  • "This standard was revised in 2003"
  • "and again in 2007"
  • "when it was split into two parts"
And that is from just two sentences from one section of the article. Other sentences and other sections have similar problems. That is why I bannered the article yesterday.
Despite not satisfying the requests for references, you removed the banner yesterday, so to help you find the problems, I tagged a few of them, individually, today. Now you have removed those tags too, and still haven't satisfied the requirement for refrences!
Canepa (talk) 08:32, 8 August 2012 (UTC)

Hi Martinvl. Here we go again. GaramondLethe 18:08, 8 August 2012 (UTC)

Arbitrary precision arithmetic?

I found this to be a little confusing:

"It is often not practical to perform these calculations directly on a modern PC..."

I eventually figured out that the limitation isn't modern PCs, but languages that don't implement arbitrary-precision arithmetic. Python (the most common example) eats this stuff for breakfast. Using the example from the article:

>>> 3214282912345698765432161182 % 97

Perhaps change this to "Is is often not practical to perform these calculations directly in without arbitrary-precision arithmetic..."?

If that's getting too technical, perhaps: "It is often not practical to perform these calculations directly..."

GaramondLethe 00:05, 9 August 2012 (UTC)

I have reworded the section. I am not worried about it getting too technical - the artcile has been designed so that the user can skip the technical sections. Martinvl (talk) 06:19, 9 August 2012 (UTC)
Your wording was better than mine. Thanks! GaramondLethe 10:30, 9 August 2012 (UTC)

Triomio edits

Hi Triomio.

Why are numeric-only codes significant? Why is a column needed to assist the reader with the task of dividing the length by four? Thanks. GaramondLethe 01:08, 10 August 2012 (UTC)

International Bank Account Number is an international standard

The intro says so, but I strongly doubt that "International Bank Account Number is an international standard". Triomio (talk) 01:33, 10 August 2012 (UTC)

ISO = International Standards Organisation" Martinvl (talk) 03:56, 10 August 2012 (UTC)
Point of pedantry:

Even the name of the organization is standardized. The name, "ISO" is not an acronym but was derived from the Greek word "isos" meaning "equal". (The relation to standards is that if two objects meet the same standard, they should be equal.) This name eliminates any confusion that could result from the translation of "International Organization For Standardization" into different languages which would lead to different acronyms.

Not sure if you consider the source to be reliable, but it's the best I can do with hotel wireless. Snide remarks concerning the elimination of confusion by introducing acronyms that appear to be either wrong or translated from the French may be directed to the Chair of Abbreviated Symbols, Department of Weights and Measures, Unseen University....GaramondLethe 04:57, 10 August 2012 (UTC)
Re: the introductory sentence: "The International Bank Account Number (IBAN) is an international standard for identifying bank accounts across national borders with a minimal risk of propagating transcription errors."
The first part of that introductory sentence is factually incorrect. The International Bank Account Number is not an international standard. What the IBAN is, is a bank account number complying with an internalional standard. There is a difference. I suggest replacing the first sentence with: "An International Bank Account Number (or IBAN) is a bank account number which complies with the format defined in an international standard. Compliance with the format ensures uniqueness of bank account numbers across national borders and provides a mechanism to identify transcription errors."
Alternatively, if the article was renamed to: "International bank account numbering standard", the sentence would only need a minor change to: "The international bank account mumbering standard is an international standard for identifying bank accounts across national borders with a minimal risk of propagating transcription errors." Canepa (talk) 08:31, 10 August 2012 (UTC)
The SWIFT registrar's "National IBAN format Registration Form" states "The International Bank Account Number (IBAN) is an international standard (ISO13616) for identifying bank accounts". A quick google shows several other sites where IBAN is referred to as an international standard. It looks like you're making a wp:truth argument, and reliable sources trump wp:truth. GaramondLethe 14:20, 10 August 2012 (UTC)
No I'm not. The definition that I gave is also reliably sourced. The abstract for ISO 13616:2007 itself gives us: "ISO 13616-1:2007 specifies the elements of an international bank account number (IBAN)" and "The IBAN is designed for automated processing" and does not suggest that "IBAN" is a synonym for the standard itself. A quick google also shows many other sites where IBAN is referred to as the resultant account number, rather than as the standard itself; try search on "An IBAN or International Bank Account Number is your bank account number in a standard, internationally recognised format.", for example.
It makes sense to describe it as an account number (the clue is in its name) and explain that it conforms to the international standard. We need to weigh "RS && semantic integrity" against "RS but illogical". Canepa (talk) 16:50, 10 August 2012 (UTC)
Well, I'm pleased that we appear to agree on the problem. I do think this is an issue of semantic integrity versus illogicality, and I'll !vote for illogicality (and its attendant clarity) every day of the week and twice on alternate Thursdays. Yes, IBAN is type of account number, and it's a standard for account numbers, and it's a particular ISO document. Semantically that's a hairball, but it's a hairball that humans have little difficulty processing. Untangling that hairball for the sake of semantic aesthetics might make for a more logically rigorous article, but it will also result in one that's much harder to read and far less useful. (This also explains why programmers prefer sloppy, illogical languages like C rather than logically rigorous approaches like Denotational semantics. When it comes to utility, expressiveness trumps purity.)
If you can convince me there's a problem here that needs to be solved, great, we can start discussing solutions. But a lack of semantic integrity (to the extent that clarity isn't compromised) isn't a problem; it's a feature.
GaramondLethe 17:41, 10 August 2012 (UTC)

Any source that claims IBAN is a standard deviates from the ISO website, and from the four ISO standards that contain the character sequence IBAN in the name. Regarding your claim about IBAN and it's a particular ISO document - it is not: - check "withdrawn". Triomio (talk) 20:53, 10 August 2012 (UTC)

I searched for IBAN at the ISO site and the standards came right up. I honestly don't know what the point is you're trying to make here. Try again? GaramondLethe 21:48, 10 August 2012 (UTC)

The article is horror

Format and validation information is spread between different sections.

Claims are made all over the article that cannot be found in the referenced documents.

"The banks of most countries in Europe use an IBAN identifier for their accounts as well as the nationally recognised identifiers" - Where is this claim backed at ????

Triomio (talk) 02:36, 10 August 2012 (UTC)

To answer your last question, I understand that this kind of legal document can be difficult to read if you don't have the requisite training. We're fortunate in that reliable sources are not required to be grokkable by all editors. I've skimmed enough of this document to satisfy myself that there isn't a problem with the citation. If you think there is a problem, your time would be better spent finding another citation rather than demanding an explanation of this one. GaramondLethe 05:27, 10 August 2012 (UTC)
Not a problem with the citation, but with the claim in the WP article. Which of the articles of the Regulation backs "The banks of most countries in Europe use an IBAN identifier"? Triomio (talk) 20:34, 10 August 2012 (UTC)
Triomio, I think you may need to illustrate your point by giving examples of some specific cases. Canepa (talk) 08:39, 10 August 2012 (UTC)
Look at this version before I edited the article yesterday. Format information is spread between the sections
  • 1.1 Practicalities - 1) The IBAN should not contain spaces when transmitted electronically. However, when printed on paper, the IBAN is expressed in groups of four characters separated by a single space, the last group being of variable length as shown in the example below 2) Table with IBAN formatting examples 3) The characters that may be used in an IBAN are the numbers '0' to '9' and the 26 upper case Latin alphabetic characters 'A' to 'Z'. This applies even in countries such as Greece, Saudi Arabia and Israel (see above) and others where these characters and/or numerals are not used in the national language.
  • 4.2 Valid IBANs by country - a table with all the formats (BTW, why is this called "Valid IBANs"?)
Features are spread between
  • 1.1 Practicalities - The check digits enable the sending bank (or its customer) to perform a sanity check
  • 2 Features - One of the design aims of the IBAN was to enable as much validation as possible to be done at the point of data entry
Examples for unreferenced claims can be found in the article history by looking for {cn}. Triomio (talk) 21:06, 10 August 2012 (UTC)

Truth or fiction - allowed characters

"The characters that may be used in an IBAN are the numbers '0' to '9' and the 26 upper case Latin alphabetic characters 'A' to 'Z'." -- Any reliable source? Triomio (talk) 03:00, 10 August 2012 (UTC)

I took a look and satisfied myself that the existing citations cover this more than adequately. GaramondLethe 05:18, 10 August 2012 (UTC)
Add references to the article! Triomio (talk) 20:30, 10 August 2012 (UTC)


Seems like the term "modulus" is not used correctly in the article. For example, there is a section titled Calculating the modulus of a large number. In that section the large number is and the quantity that is calculated is , which isn't the modulus. The modulus is . The number that is calculated is the remainder after division of by the modulus . --Bob K31416 (talk) 03:03, 10 August 2012 (UTC)

Hi Bob. Based on this Wolfram page I think you're right in the strict math sense, and I'd like to see the article be made... well, congruent with that sense if doing so won't confuse the C programmers. The correct technical term is "common residue", right? I don't think that's helpful. Is the better solution to rename the section "Using modular arithmetic with large numbers"? GaramondLethe 04:45, 10 August 2012 (UTC)
Re "The correct technical term is "common residue", right?" — It is a correct technical term for the concept. Another would be remainder. (See also Modular_arithmetic#Remainders.) In any case, I think the term modulo operation would be better to use than "modulus", which seems to be wrong. For example, the section title could be changed to "Performing the modulo operation on a large number". --Bob K31416 (talk) 12:49, 10 August 2012 (UTC)
Bob, now that things have calmed down a bit I went ahead and put in something similar to what you suggested. "Modular arithmetic" is a little more common than "modulo operation", but I don't have a strong preference either way. Your thoughts? GaramondLethe 04:41, 11 August 2012 (UTC)
I prefer modulo operation because it is more specific and there is a good explanation of it at the beginning of its Wikipedia article. Note: 1) that the purpose of the section is to support item #6 that precedes it, i.e. "Calculate Mod-97 of the new number"; 2) that item #7, "Subtract the remainder from 98 and..." uses the term "remainder" for what is going to be calculated. It looks like some rewrite is needed to coordinate the terminology regarding "modulo operation", "Mod-97", and "remainder". Also, "modulus" needs to be changed in the section title "Example of modulus calculation". --Bob K31416 (talk) 10:39, 11 August 2012 (UTC)
Modulo operation does indeed have its own entry separate from Modular arithmetic. Want to go ahead and make a pass through the article? GaramondLethe 15:54, 11 August 2012 (UTC)
Thanks. I made a pass through the article.[4][5] --Bob K31416 (talk) 16:46, 11 August 2012 (UTC)
I took a look and that's definitely an improvement. Thanks! GaramondLethe 17:33, 11 August 2012 (UTC)
Some more potential improvements for anyone to consider doing: 1) Use ISO 7064 for source instead of computer code; 2) replace non-standard notation —  ;   etc. --Bob K31416 (talk) 17:56, 11 August 2012 (UTC)
I would just write: D mod k in this case. Do you have something else in mind? Tomeasy T C 01:10, 12 August 2012 (UTC)
Just a moment ago I made that edit re just parentheses, and similar ones. --Bob K31416 (talk) 02:55, 12 August 2012 (UTC)

Triomio's changes 9-10 August 2012

I have undone all of Triomio's changes that were made on 9-10 August 2012. This article is the cummulation of many years work by a number of editors, so if Triomio wants to be helpful, (s)he can do so by finding citations, not by asking others to do it. Also, the lede paragraph does not have citations - the relevant citations are all in th ebody of the article.

Also, if you are unhappy about things, this artcile is downloaded by many banks daily, so please do changes one at a time so that they can be properly discussed. Martinvl (talk) 04:04, 10 August 2012 (UTC)

Triomio went to a lot of trouble describing each of the changes made. You should answer each point individually, not apply a curtly described blanket revert. For example: which specific citation do you believe supports the first part of the first sentence: "The International Bank Account Number (IBAN) is an international standard..."?
Can you substantiate your claim that "this artcile is downloaded by many banks daily". Canepa (talk) 08:47, 10 August 2012 (UTC)
If you go to the Page View Statistics, you will see that this page gets between 3k and 4k hits Monday to Friday and about 1k hits Saturday & Sunday. Now copy the first paragraph of the article intio the Google search box and see what you get. Verbatim copies of the lede paragraph on lots of websites belongiong to banks. That is why I believe that this article is downloaded by many banks daily. Maybe you have an alternative explanation. Martinvl (talk) 12:42, 10 August 2012 (UTC)

Martinvl, most of my changes where one at a time - it is exactly you who makes broad changes to the article, wholesale insertion of nonsense and removal of requests for citations. I don't see why an article that is "downloaded by banks" should be treated different to other articles. Maybe the 3k-4k hits come from pupils from schools? And they read all the nonsense. Stop presenting nonsense! Triomio (talk) 20:40, 10 August 2012 (UTC)

Removal of requests for citation

Why are requests for citation removed? E.g. by User Garamond Lethe:

Triomio (talk) 21:45, 10 August 2012 (UTC)

Because the relevant citations are in the article body. Usually we don't use citations in the lead section because it only sums up what has been written further down the page, and that is where to cite stuff. See Wikipedia:Citing sources#When and why to cite sources. De728631 (talk) 21:54, 10 August 2012 (UTC)
Triomio, please self-revert your last batch of edits. I think you've blown past WP:3RR and you may be looking at a temporary block for edit warring. GaramondLethe 22:10, 10 August 2012 (UTC)

At De728631 - the relevant citations are not in the body, they are nowhere. Triomio (talk) 22:21, 10 August 2012 (UTC)

Having read much of the text of the first two citations, I'm honestly not sure what you're talking about. Can you give me one example? GaramondLethe 00:22, 12 August 2012 (UTC)

Removal of sourced terminology

User Garamond Lethe with this edit:

turned "IBAN print format example [1]" into "IBAN formatting example". The examples only cover the IBAN print format, which is named so in the swift document. So what? Triomio (talk) 22:21, 10 August 2012 (UTC)

I think you've answered your own question. As there is only a single format (print), correct English does not add a qualifier. ("print format example" implies the existence of a "non-print" format example.) GaramondLethe 00:18, 12 August 2012 (UTC)
As I understand, there is the common printing format and the format for electronic transaction (non-print). They differ in having/ not having packages of 4 characters separated by single spaces. Tomeasy T C 01:07, 12 August 2012 (UTC)

Reflist for above ref.

--Bob K31416 (talk) 19:11, 17 August 2012 (UTC)

Changes 13 August 2012

User:Canepa made some observations regarding the use of the IBAN. I have reworded sections to state that the variouas countries have implemented the IBAN - this is the exact word used in the relevant IBAN document. I do not understand assertion that the statement "...been adopted by other European and Middle Eastern countries" is "not supported in the main article body" - Israel, Saudi Arabia, Palestine and Kuwait are in the Middle East. Martinvl (talk) 13:31, 13 August 2012 (UTC)