Report on theEuro - - in IT Standardization
This report was approved on 1999-04-23 by CEN TC304 [European Localization Requirements] at its Plenary in Tübingen. It has been preceded by two interim reports (of 1998-06-30 and 1998-10-12) and a draft of 1999-02-11 for the final report (circulated as TC304 N881).
In this version, vendor-specific information is given in the form of references to their web sites in Appendix B.
The Currency Code for the euro for use in monetary transfers by financial institutions et al. has been registered in the three-letter form as EUR and in the numeric form as 978 with the international registration authority BSI [British Standards Institute], appointed by ISO according to standard ISO 4217. Note: This three-letter code does not follow the normal rule: the two-letter country code (EU, not eligible for registration as such, although reserved) followed by the first letter in the name of the currency (which normally would be E).
The euro sign has evolved from the initial logo to also a new character. The logo is defined as a rounded E with double line for the central horizontal bar, blue with yellow background (ref: COM(97) 418fin). The logo may be downloaded from web site http://europa.eu.int/euro/.
As a character, the shape of the euro sign needs to adapt to the textual environment. For this purpose, the euro glyph has been registered as ISO Glyph ID 8059 in the International Glyph Register, Volume 1: Alphabetic Scripts and Symbols with the international registration authority AFII [Association for Font Information Interchange], appointed by ISO according to standard ISO/IEC 10036; the registration authority is currently being transferred to UNICODE.
As the shapes of the glyphs are adapted to the respective fonts' typographic properties, they are generally narrower than the logotype shape.
Proposed alternative shapes for an OCR-B euro glyph have been developed within the Euro Project Team and reported in a CEN TC304 document, as discussed later. Further standardization work is needed in this area.
As a character, one must be able to input the euro sign to an IT system (using e.g. a keyboard), to process and store it, to transmit it, and to output it (to e.g. a display, printer, et al.). Most of this requires that the sign be included in the coded character set of the IT system in question.
The support for the euro sign as a coded character in the various environments is well on its way.
The inclusion of the euro sign in the multi-octet Universal Character Set (ISO/IEC 10646-1 AM 18, and UNICODE) has been finalised by ISO/IEC JTC1/SC2 [Coded Character Sets]: The normative name is EURO SIGN and its 16-bit code is 20AC. This has been approved as an International Standard in April 1999. Note: EURO SIGN is not to be confused with the basically unused EURO CURRENCY SIGN (20A0) in the same standard(s).
The 8-bit ISO/IEC 8859-15, "Latin-9" that includes the euro sign has been finalised: The code for EURO SIGN is A4. This has been published as an International Standard on 1999-03-15.
In the context of "Latin-9", SC2/WG3 also recommended (Resolution M 13.08) that the equivalent position used by the International Currency sign (ICS) be used for the assignment of the Euro sign in new graphic character set registrations or standards (that are based upon existing 8-bit standards where the ICS is currently encoded).
The revision of the 8-bit ISO/IEC 8859-7, "Latin/Greek" is planned to include the euro sign. The draft presented in the September 1998 meeting of SC2/WG3 indicates that the code for EURO SIGN will be A4.
A new work item proposal was approved by CEN TC304 in November 1998 to revise EN 1923 (European Character Repertoires and Their Coding) to incorporate the euro sign into it, and additional guidance to this work was given in April 1999. Other coded character set standards may be created or extended to include the euro sign. For certain old standards, e.g. ISO 6937 (Coded Graphic Character Set for Text Communication Latin Alphabet) and all 7-bit coded character sets, a decision not to extend has been made.
The euro sign will be included in the forthcoming Multilingual European Subset repertoires of ISO/IEC 10646 (MES, which have been initiated by a CEN TC304 Project Team but are now subject to a CEN/ISSS Workshop and, if approved, will have the status of CWA), which will also be the base for e.g. the European Ordering Rules, two ENVs to be produced by other TC304 PTs. If the MES CWA cannot be agreed upon, TC304 will consider amending ENV 1973 (Minimum European Subset of ISO/IEC 10646) with the euro sign; otherwise, ENV 1973 will be withdrawn.
There will be several industry or de facto standard implementations (for both workstations and mainframes) available shortly if not yet. Vendor provided web site addresses giving information on these implementations are referred to in Appendix B.
In line with the above ISO/IEC JTC1/SC2/WG3 recommendation on the Euro sign vis-à-vis the International Currency sign, the Swedish IT Standardization has applied for ISO registration of three code tables with the euro sign included in this way. These tables correspond to the "right halves" of the ISO/IEC 8859-1, -4, and -13 (aka "Latin-1", "Latin-4", and "Latin-7"). The tables were registered in September 1998, and can now be referred to as ISO-IR 204, 205, and 206, respectively. For IT applications needing the euro sign, it is therefore possible to formally specify the use of such alternatives as ISO-IR 6+205, corresponding to Latin-4, etc., when an applicable international standard is not available or cannot be used.
In HTML (HyperText Markup Language) and XML (eXtensible Markup Language), the euro sign is referred to by its hexadecimal or decimal UCS code, i.e. € or €. In HTML Version 4, it can also be referred to as €.
Where the euro sign is not available by any of the above means in the applicable coded character set or its presentation, it is recommended that, when there is room for expansion, the euro sign be replaced by the three letter currency code EUR. When there is no room for expansion, a capital E would be the next level of fallback, unless it would create problems in its context. Ultimately, a user specific fallback may need to be defined.
In this report, the primary input mechanism to be described is the keyboard, although scanning and related OCR-B and bar codes are also discussed.
In some applications such as word processors, the editor program permits insertion of symbols without the need to have a keyboard position. In these vendor specific applications, the user can enter the euro sign into the text by clipping and dropping from the symbol palette as long as the symbol is available in the font supported.
Voice recognition has not been discussed at all, nor have the special needs of input by handicapped people, to whom the euro sign is but one of many details. For the web, the Web Accessibility Initiative (WAI) at the World Wide Web Consortium (W3C) addresses accessibility on many levels, in partnership with other organizations which are committed to ensuring that this new technology is accessible to all, including people with disabilities.
The keyboards can be divided into general purpose and special purpose keyboards.
On the keyboards, there is no international standard in existence or in process that would cover the national primary keyboard layouts in its scope. Also, in most EMU countries, there is no national keyboard layout standard in use, although some exist.
The EU Commission has produced a recommendation which it is likely to enforce for the procurement for the EU / EEA official business. The European subsidiarity principle is very strong and it would prohibit CEC from mandating anything for the national keyboard layouts. The non-existence of any such international (world-wide or European) standard is essentially for the same reason.
The Open Workshop on the Euro in IT Standardization held in Brussels in February 1998 decided to support the CEC recommendation to use AltGr+e for most European countries (on keyboards where Alternate Graphics is available). After the Euro workshop, CEN TC304 has passed a similar resolution. Thus, CEN TC304 is the only wide area standardization body that has made any recommendation for the placement of the euro sign in the primary layout.
The ISO 9995-3 standard for the common secondary keyboard layout is only intended for use when the national or local primary keyboard layout, as specified by the appropriate National Standards Body or e.g. a vendor, does not include the euro sign. The recommendation in that standard is to place the euro sign as an additional character on the fifth key from the left on the top row (position E04). One should note, however, that there is no standardized method to access this secondary keyboard layout (which frequently is not implemented at all), although some work in this area has been initiated in ISO/IEC JTC1/SC35 [User Interfaces].
In line with the European recommendation, several vendors will support keyboards where AltGr+e is used for the euro sign in most European countries; in addition, for certain countries, some will also support a second location. Incidentally, the support does not necessarily require a new keyboard, normally a new keyboard driver will do, which is the primary rationale for the recommendation: e for euro.
National standardization on keyboards cannot be expected to provide timely guidance to industry on where to place the euro sign on general purpose keyboards. It seems that the PT on Keyboards that CEN TC304 has set up, can in fact play a significant role in the revival of keyboard standardization in Europe.
Note: For other than workstation keyboards, a harmonised solution does not appear to be possible.
The placement of the euro sign on special purpose keyboards (e.g. on cash dispensers, payment terminals and cash registers) is still wide open and can only be addressed by the relevant industries and the related standardization bodies. In several instances, alternate means of specifying the currency already exist, e.g. by icon selection.
In a more general field of application, CEN TC304 has asked ETSI (or its User Group together with ANEC [European Association for the Co-ordination of Consumer Representation in Standardization]) to come up with a recommendation for the placement of the euro sign on telephone keypads (particularly on mobile phones), which also has implications on their character encoding.
Scanning of textual documents relates to the subject of the ISO/IEC JTC1 Business Team on Imaging among others.
The relevant OCR-B standard is ISO 1073/II-1976.
A new work item proposal was approved by CEN TC304 in June 1998 for an EN to add the glyphs for the euro sign and possibly a number of other European language characters to the glyph repertoire of OCR-B, as specified in ISO 1073/II-1976. The glyph for EURO SIGN will be normative. The work for adding the euro sign to the repertoire of the OCR-B standard is within the scope of the Euro PT.
As the result, one could assume that the requirements for reliable, high speed scanning could be met for e.g. the purpose of sorting documents using the euro symbol in connection to monetary amounts. At the same time, the requirements for reliable document archive can be met, as OCR-B supports the most reliable scanning method based on current technology. There is also other interest in defining the euro sign in this context.
After development of alternative shapes for the OCR-B euro glyph, a draft report was produced (TC304 N837) documenting the issue. This report was posted for possible comments also on the CEN/ISSS web site. In addition, the creator of the original OCR-B font, the typographical designer Adrian Frutiger, was consulted in the matter.
The two alternate OCR-B glyph shapes developed will be the subjects of testing in the new Project Team, in close collaboration with the Euro PT. The Call for Experts for the testing team was issued in February 1999. Meanwhile, work will proceed within the Euro Project Team on a draft for a new OCR-B EN or ENV, based on the ISO standard 1073/II, and including the euro sign in a tentative shape.
The bar coding standards are the subject of ISO/IEC JTC1/SC31 [Automatic Identification and Data Capture Techniques] and CEN TC225 [Bar Coding].
Of the bar coding standards, EN 797 ("EAN/UPC") only specifies numbers 0-9.
When using EN 799 ("Code 128") that supports 8-bit coded character sets, the default set is ISO/IEC 8859-1. By mutual agreement between the parties, also other coded character sets may be used, e.g. ISO/IEC 8859-15 (or 8859-7) or ISO-IR 6 together with ISO-IR 204, 205 or 206, all of which support the euro sign in position A4.
When using EN 800, also ISO 16388 ("Code 39"), which only supports the 7-bit ASCII character set, the euro sign can be represented by the numeric (978) or alphabetic (EUR) currency code according to ISO 4217.
A new bar coding standardization initiative, "Ultracode", which has not yet been formally proposed as a new work item, is planned to be based on ISO/IEC 10646-1 / UNICODE, in which case it would include the euro sign in position 20AC.
The display and printer output is the only one to be discussed presently. For scannable output, many of the issues are the same as those for input.
Audio output has not been discussed at all, nor have the needs of handicapped people (e.g. Braille). For Braille, the relevant committee is ISO TC173 [Technical Systems and Aids for Disabled or Handicapped Persons] and the relevant standard is ISO 11548, which deals with 8-dot presentations. The Braille presentations, particularly with the 6-dot basic Braille, are even more language and culture dependent than coded graphic character sets ever, which has an impact on any standardization effort related to them.
Several font manufacturers already provide the euro sign in their font collections.
The display manufacturers are not expected to face any particular problems in supporting the euro sign, although, depending on the implementation, various levels of software are likely to be affected.
Many printer manufacturers are adding the fonts for the euro sign to their downloadable fonts. In some instances, the system fonts can be used to supplement the built-in fonts; the latter will take more time to replace (and may only be extended for new printers). The mechanical printers, such as the daisy wheel printers, will require new hardware parts plus changes to the software needed to support them; the high speed band printers (which are primarily used for the back office operations at major financial institutions) are not likely to change in the near future.
The presentation of amounts in euro has several aspects to be solved.
Each country has currently their own preferred convention to display amounts in their national currency. This national individualism is expected to extend to the display of amounts in euro. Several countries, however, do not have conventions that can be readily carried over, as e.g. the magnitudes of the unit values can be significantly different.
The question of national preferences is being addressed by the CEN TC304 Project Team on Populating the Cultural Registry. A summary report on the presentation of monetary values, particularly in euro, will be made available on the TC304 web site http://www.stri.is/tc304.
In spite of the cultural differences that lead to the individual national preferences, a common way to display amounts in euro is required for conducting the official, pan-European business of EU, EEA, CEN, and related organizations, in order to provide for identical, unambiguous understanding in these multicultural European environments. The same is required for many unrelated multinational applications, e.g. for the commercials on the satellite channels.
The CEN/ISSS Workshop on European Default Locale is addressing this issue. This CWA has had some unrelated external dependencies (on the MES CWA), which are in the process of being resolved for it to be finally agreed upon. Nevertheless, the current draft version of Clause 4: Monetary formatting, is given in Appendix A.
As the current locale structure in many IT systems (notably POSIX) does not allow for multiple, different occurrences for the locales dealing with the presentation of monetary values, the need for the IT vendors to give the users the right default values for the concurrent display of amounts in their traditional currency and the euro, will require some quick solutions. Contributions have been made primarily to ISO/IEC JTC1/SC22/WG20 [Programming Languages and Software Interfaces; Internationalization] in support of dual rendering under application control. Some involve changes to the locale structure and the corresponding calling APIs, and others involve calling two separate locales using modifiers as the means to select the appropriate locale without changing the Application Programming Interface. The latter seems to be the more commonly favoured approach.
In connection to the introduction of the euro sign into IT systems, a number of interoperability issues need to be addressed. Many of these are related to the fact that the different 7- or 8-bit coded character sets are indeed different. Until the UCS / UNICODE will be uniformly supported and used, user systems with different coded character sets may encounter interchange problems - and certainly will if they are not prepared for the possibility.
The introduction of the euro sign into the standardized coded character sets has been done with the intent of causing the minimum level of disruption.
In the case of ISO/IEC 10646-1, the new code is an addition to the old codes, all of which continue to be used as such in the new environment. If the target system does not recognise the code for the euro sign, it knows that it has received an unknown character and it can identify it to the users as such.
In most 8-bit environments, all bit combinations have already been used to represent some graphic or control character. Thus, the introduction of any new sign commonly requires that some other character be removed from the repertoire.
In the case of ISO/IEC 8859-15 ("Latin-9"), although the new coded character set has considerable resemblance to ISO/IEC 8859-1 ("Latin-1"), a number of characters available in "Latin-1" are not available in "Latin-9" and vice versa. If the user is not aware of the character set that is used to encode the data he is receiving, quite some misunderstandings will be caused, as the bit combinations all look alike, be they used for 8859-1, -2, -3, or whatever. In the case of "Latin-9" vs. "Latin-1", the severity of the potential misunderstandings has been lessened by carefully selecting for the differences only positions that have not been widely used in "Latin-1". In fact, it is relatively easy to check the databases, whether any of these confusing characters are currently in use; if not, one can relatively safely assume that future use is for the characters of "Latin-9".
Another cause of problems may be tagging (of e.g. data base content, transactions, and e-mail), which, on the other hand, is actually the way for the future. An example of a problem situation is when the originator would tag a transaction as using a new coded character set, which set is not (yet) supported/recognised by the target data base system. If the target system is instructed to reject any transactions with unknown coding as invalid, the data is not accepted even if it is known that no discrepancies would result. In cases like this, some adjustments ("faking") may be needed during the transition period at either the sender or the receiver or both.
A specific elaboration report on this is being prepared for publication on the CEN TC304 web site.
Other aspects will be addressed as they arise. Although the Euro Project Team will conclude its work, the subject falls into the area of CEN TC304.
Erkki I. Kolehmainen <firstname.lastname@example.org>
TC304 Project Team Leader for the Co-ordination of the Euro in IT Standardization
Draft for European Default Locale Clause 4: Monetary formatting
This default is for general use, and not applicable for use in value documents.
The name of the currency is "euro", subdivided into 100 "cent".
The grouping of large monetary amounts is in groups of 3 digits.
The group (thousands) separator for monetary amounts is FULL STOP <.>.
The decimal separator for monetary amounts is COMMA <,>.
A positive monetary amount has no sign attached to it.
A negative monetary amount is preceded by a HYPHEN-MINUS <-> in front of the currency symbol.
The 3-letter currency code according to ISO 4217 is: EUR
The currency symbol is:
Both the 3-letter currency code and the currency symbol are placed in front of the monetary amount and one NO-BREAK SPACE separates the currency code and the monetary amount, while no space comes between the currency symbol and the amount.
Examples of formatting of monetary amounts:
3-letter ISO 4217 currency code: EUR 543,21
(Domestic) Currency symbol: 543,21
Use of negative sign: -543,21 -EUR 543,21
Thousands and decimal separators: 9.876.543,21 EUR 9.876.543,21
NOTE 1: In the case that the EURO SIGN cannot be used in an environment, because for example of hardware or software limitations, it is recommended to use the 3-letter ISO 4217 code "EUR" instead, or the capital letter "E" where only one character is permitted.
NOTE 2: When other currencies are to be displayed, the formatting specified here for the 3-letter currency code or the currency symbol for the Euro currency is also used, using the normal decimal precision where applicable.
NOTE 3: The 3-letter ISO 4217 currency code is called "the international currency symbol" in the ISO/IEC 9945 POSIX standards.
Links to vendor information web sites
Note: These web sites are those of which the PT has been informed, usually by active participants in the Euro project. While efforts have been made to verify the accuracy of this information, no guarantee is given that the referred web sites are operational or provide the expected information. Many vendors provide also product-specific pages (including fixes for downloading) that can be reached via the addresses given.
International Business Machines (IBM)
http://www.ibm.com/ (search for euro)