Notes from the 3rd Euro-Workshop arranged by CEN/TC304 3 Feb 1998

The 3rd Euro Workshop was held at CEN/ISSS in Brussels 3 February 1998. The previous day the plenary of CEN/TC304 had approved the selections of experts for the Euro Project Team. The Workshop was held to collect input for the work of this Project Team.

These notes are not the formal minutes of the meeting as there will be no formal appoval of them. They are, however, jointly agreed to by the Chair and the Secretary.

The agenda was as follows:

1 Opening of the meeting
  Chair: Erkki I. Kolehmainen, Secretary: Þorgeir Sigurðsson.

2 Roll call of participants

3 Adoption of the agenda and identification of documents;

4 Chair's prerogative:
 - Euro sign: from logo to character symbol.
 - Need to be able to
   -- enter from various input devices (notably keyboard, also optical &al.),
   -- process and store in an information system and interchange between them,
   -- produce in a human and machine readable or audible output form.

5 Status of / activities for Registrations:

 - Currency code (EUR) for financial transfers.
 - Logo?
 - Glyph to facilitate use as a character symbol.
 - As a cultural element.

6 Input/Output

 - Keyboard:
    WW (ISO) stand on kb layouts.
    Need for a European (CEN) standard for kb layouts?
    Recommendations for national standards for kb layouts.
    Recommendations/implementations for CEC and related use.
    Implementation plans/status by the IT industry.
 - OCR-B.
    Scanners
 Output:

7 Electronic code for information interchange:
 16-bit UCS (10646 and its various subsets, MES/EES), UNICODE
  - Preparation for the disposition of NB comments on the SC2 proposal on 10646-1.
 8-bit coded character sets
  - Preparation for the disposition of NB comments on the SC2 proposal on 8859-15 ("Latin-0").
 Other proposals for the coding of Euro.
 Interoperability.

8 Other aspects:
 sorting, matching,
 writing of amounts, simultaneous rendering.

9 Report on national Euro projects.

10 Conclusions; PT meeting to follow.

----------------------------------

1 Opening of the meeting

Erkki Kolehmainen welcomed the participants and said that he hoped this would be a successful meeting and this would be the last open Workshop meeting to gather input for the CEN/TC304 Project Team on the Euro.

2 Roll call of participants

Those present included: Erkki Kolehmainen (Finnish Data Communication Association), Wolf Arfvidsson (Chair of CEN/TC304 also a chair of a Swedish national committee for the introduction of the Euro), Johan van Wingen (Chairman of a Dutch national committee on coded character sets), Wilmart Jacques, (Manager of IT Standards, Siemens Nixdorf), Tom Garland (Sun Microsystems, i18n/l12n)), Chris Makemson (UK), Klaas Ruppel (Finland), Evangelos Melagrakis (Convener of ISO/IEC JTC1 SC2/WG3 8 bit coded character sets), Keld Simonsen (Denmark, editor of ISO 14652 Cultural Conventions), James Boyd (CEN/ISSS), Þorgeir Sigurðsson (secr. of CEN/TC304), Adele Folletti (CEN/ISSS Workshop manager), Mike Ksar (HP, 10646-SC2/WG3 convener), Arnold Winkler (Unisys, i18n- SC22/WG20 convener), Karl-Ivar Larson (consultant,ex Ericson), Knud Nordby (Telia, ETS te HF), Monica Ståhl (IBM Sweden), Wera Lundström (ITS Sweden), Kenneth Thompson (DGIII, European Commission), Graham Dixon (UK), L. Tossounidis (EC - Informatics Directorates), Borka Jerman Plazic (Slovenia)

3 Adoption of the agenda and identification of documents

The agenda was adopted with the changes that have been included in these notes. Erkki made the change that Œoutput¹ was missing from agenda item 6 and Mike Ksar added scanners to the list of output devices.

Erkki listed documents that were tabled at the meeting:
Draft agenda (N789)
Euro Recomendations from the second Euro Workshop Workshop (N771)
Document from Keld Simonson on the Euro in ISO/IEC 14652 (ISO /IEC JTC1/SC22/WG20 N553)
Text of email by Karl-Ivar Larson about inclusion of Euro sign in OCR-B.
Notes on inclusion of Euro sign in 8-bit coded character sets, by Karl-Ivar Larson.
Expert contribution from Klaas Ruppel (Note: This is a National body contribution in N794)
Documentation from AFII on the registration of the Euro Glyph.

4 Chair's prerogative:

Erkki Kolehmainen gave a speech to set the scene. He said that even if he would chair the meeting impartially, he wanted to make it clear, that he was in full support of the Euro Sign, including full support for the Euros Sign in IT standards.

Erkki gave some background information on the Euro Sign. He said that it was originally designed as a Logo, for specific use. When it became a character, its nature became different. Erkki said that last summer, there had been heated discussion in standardization circles on the need for the Euro Sig, but the discussion was different now. As an example for the need, he said that when he saw newspaper articles on currencies, he saw the Dollar and Yen a Signs and would expect to see the Euro Sign together with them. - He said IT standardisation now needed to give the Euro Sign all the capabilities that a character needs to have, for it to become a fully fledged currency symbol.

A discussion resulted from this on the general use of currency signs versus three letter codes. Erkki said there had been a political decision to introduce the Euro sign and it was our duty to facilitate its use. The use of the Euro Sign as a character in text would not conflict with the use of the code ŒEUR¹ for financial transactions. Chris Makemson pointed at the British usage of the pound sign and said that even if most computer systems use three letter codes, there were some that used only one digit. Van Wingen gave his personal opinion (he said this had not been discussed in NNI) and said that there used to be a Dutch currency sign, but it was decided it was not needed and that for the moment a small f was used. He said that even the dollar sign was ambiguous and at the moment there was an increased use of three letter codes. Borka told of her experience in Amsterdam where three letter codes were not used on train tickets. Erkki concluded by saying that standards should allow competition between different ways of expressing currency values.

Keld Simonson wanted to clarify that it was not just the Euro Sign but also other aspects of displaying monetary values that needed to be addressed. There was some discussion on the placement of the Euro Sign in monetary amounts. Keld Simonsen said this would be addressed in the Project Team working on the European default locale.

Karl-Ivar pointed out that in high-security applications, such as using optical recognition in banking there was no room for competing approaches. The shape of the Euro Sign and its use needed to be specified accurately.

Erkki asked Ken Tompson from the European Commission on his position. Kenneth Thompson said that the Commission had asked for the Euro Sign and it had asked CEN to assist with its introduction in IT standards. He said that the effort of the Commission had not been focused enough and that there was a large vacuum and guidance and proposals were more important than formal standards.

5 Status of activities for registration

Registration issues were reviewed. On Logo discussion Mike Ksar said that this would have nothing to do with AFII (a registration authority on glyphs). In discussion on intellectual property rights, Kenneth Thompson said such issues should not be our concern. Keld Simonsen said we should minute that we were not dealing with the Logo issue

The Euro Glyph was registered in October according to a document from AFII that was presented.

There was discussion on ENV12005, (a European prestandard produced by CEN/TC304 that is currently proposed for an ISO standard). Erkki said that 12005 did not presently allow for dual currencies to be displayed. He said work on 12005 was an ongoing project and this should be considered in the revision of 12005. Keld said that we could actually register dual currencies in the narrative description of ENV12005.

There was not a long discussion on registration issues, which reflected that these issues were not considered to be of immediate urgency.

6 Input/Output issues for the Euro

Keyboards

In the Agenda, the following questions were asked about keyboard: Was there a need for a European (CEN) standard for kb layouts? Should recommendations be made for national standards for kb layouts? Should recommendations/implementations for CEC and related use be made? What were the implementation plans/status of the IT industry?

There was a long discussion on these issues. Nobody from ISO/IEC JTC1 WG5 (the working group that deals with ISO keyboard standards)was present at the meeting, but Monica Staahl gave an overview of what she thought would be the position of this Working Group. Arnold Winkler said that people should be asked before new keyboard layouts were proposed and that WG5 should ask the users for their opinion. He said that in his opinion the keyboard proposal from the EC and from Microsoft were better than those produced by WG5. He added that nobody knew if we would ever get an implementation of the proposals from WG5. Lazaros Tossounidis, (the editor of the EC keyboard recomendation), clarified that it was originally the Commission that had approached Microsoft with the proposal for AltGr + e solution, it had not been the other way around. Karl-Ivar talked of the International Currency sign, its possible replacement by the Euro Sign and the effect this would have on keyboards.

Mike Ksar said that TC304 had to take a position on the keyboard proposals. He said that the Commission had put forward a proposal and that it had been accepted by industry. Any discussion on 9995-3 would take a long time and industry would not wait. He repeated what he had already said in the EuroWorkshop in October 97 that National bodies should vote against the proposals from WG5. Erkki commented that the next meeting of WG5 would be in Reykjavik and that he would be there. Mike Ksar strongly recommended that TC304 should adopt the recommendation from EC. Lazarous Tossounidis pointed out that the proposals from the EC were in two parts, one was a short term solution and another was a long term solution. Monica Staahl said that the problem with the ISO keyboard standard 9995-3 was that it was very seldom used. She said 9995-3 did not specify national keyboards and that it was not in its scope to recommend positions for national keyboards. There was a discussion on what the European proposal meant for the Irish keyboards. Lazaros Tossounidis told of an Irish keyboard layout that used dead accent to input the letter é.

Keld said that we should not recommend voting down the ISO keyboard standard, but rather help ISO WG5 to make their standard more sensible. Kenneth Thompson said we could ignore WG5 which had actually killed itself. He said in the current situation the keyboard from Lazaros Tossounidis should be accepted.

There was a discussion on the short and long term proposal from Lazaros Tossounidis. There seemed to be a consensus that even if the short term proposal was endorsed, the long term solution should be the subject of future standardisation.

Johan van Wingen said that there was a keyboard standard in the Netherlands. He said nobody was prepared to pay for its revision, which meant that it was really not relevant.

Kenneth Thompson and Erkki concluded that from the Euro Workshop there should not come a recommendation for a European keyboard standard for the Euro. The current status should be minuted, which is that the IT industry is recommending that the short term proposal of the recommendation from the Commission should be implemented. The long term recommendation of adding a new specific key should be ignored. Erkki said different solutions should be allowed to come out of free competition. Kenneth Thompson said on this that the official position of the Commission was to follow international standards and this should be the future aim, for some numbers of years the Commission would however, buy equipment according to its short term recommendation.

Erkki said there was a question if we could actually make any recommendation. Kennet Thompson said we should keep our nose out of National keyboards. Situations where multiple European languages were used, in any country, anywhere, should be our concern and it was important to specify the target for any recommendation. Lazaros Tossounidis responded to this and said there was no need to have too much respect for who could be addressed and who could not. He said guidance was urgently needed now, for everyone.

A discussion followed on important but primitive keyboards, without an AltGr key such as in banking and many financial services. Chris Makemson offered help to the Euro Project Team in getting information from this community. He said he had in a past incarnation had APAC(?) as a client of CMC (his consultancy), and another client dealing with displaying amounts on petroleum pumps. He could offer statements of problem from these two interesting clients. This was an action gratefully accepted by the chair, Erkki Kolehmainen.

OCR-B

Karl- Ivar explained what OCR-B stood for. He told of the history of the OCR-B standard. He said it was used for high security inputting of characters through scanning and had traditionally only included the characters of US ASCII. He told of his attempt of adding more European characters to this standard, but said it had failed because funding and interest for testing the specific design of the extra characters had not been available.

Karl-Ivar noted however that in the current repertoire there were currency signs (dollar, pound, yen), and that it would be natural to add the Euro Sign into this repertoire. He said that after the last meeting in Dublin (October 1997) he and Marc Küster had started investigating the possibility of adding only the Euro Sign to the OCR-B repertoire and adding the other, already defined European characters (untested) into an informative annex. Karl-Ivar had prepared such work, among other things by contacting the original designer of the OCR-B glyphs.

Evangelos Melagrakis asked for clarification about putting new characters into an informative part of the present CD. Karl-Ivar said there were several options and that one option would be for him to carry on this work without support from the Euro team. Erkki asked the audience if there was will to add the Euro Sign into OCR-B. Chris Makemson responded to this and offered two actions on himself: One was to contact his clients about whether there was a need to put Euro Sign on checks. Another action was to contact people within BSI, which had formed a special division that could handle testing of new characters in OCR-B.

Van Wingen said that it any high security situation, for which OCR-B was designed the three letter code: EUR would be used. Erkki answered and said that Chris would verify the need for an OCR-B design of the Euro glyph. Karl-Ivar said it would not be easy to verify what the needs would be, as people do not generally have need for what they have never had. Arnold Winkler said he could confirm that it was considered important to be able to scan in the Euro Sign by the scanning industry. Lazaros Tossounidis wanted to add that companies were calling him and that answers needed to be given. He gave an example from Luxembourg, where automatic changing machines must be able to deal automatically with many currencies.

Evangelos Melagrakis said that standards should include things, but then the users decided what to use. The standard should have all the characters needed and there was a basket of characters that needed to be added to OCR-B. To this Karl-Ivar replied that he agreed in principle, but he had spent a lot of time trying and there simply had been no interest for new characters in the OCR-B. He did not want to wait longer as there was little time for the introduction of the Euro. He said he had prepared a plan and that this plan needed funding, if necessary he would work for free, but the professional testing facilities did cost money.

Erkki said that he wanted the meeting to make a recommendation. He suggested stating that the Euro sign should be included in the OCR-B, and also that Karl -Ivar should submit a proposal to the March meeting of SC2 in Redmond. He could submit a proposal and then amend it in the meeting. - We should have a resolution stressing the importance of the OCR-B.

There followed a discussion on financing the OCR-B work. Chris Makemson said there were various sources for funding. Mike Ksar asked what the cost was. Karl-Ivar said that according to his plan the cost was 15 man-days as funded by the Commission. Marc Küster said it would be unfortunate if testing got financed but not the author of the proposal (Karl-Ivar). He said that people that used OCR-B applications were interested in this work. Chris Makemson said that one of his customers would be very interested in setting up the testing facilities.

Karl-Ivar concluded that he would send in the proposal to SC2.

A discussion on the Euro Sign in telephony now took place. Karl-Ivar said that one application of the OCR-B standard was that the code-table standards developed by ETSI referred to the OCR-B standard for the look of the glyphs.

Knud Nordby (chair of ETSI HF) said that this was not so important to the telephone standards because there was a limited number of dots in the array for displaying glyphs in hand held telephones and the OCR-B look of the glyphs could not be rendered accurately.

Knud Nordby noted that there was no Euro Sign on telephones. He asked for recommendation for where to put the Euro. (Characters are entered via shifting the numbers on the numeric keypad) There were some reactions to this request. Arnold Winkler said the key used to enter e was natural. Lazaors Tossounidis said that there was no EC proposal for telephone keypads. Kenneth Thompson said all the currency sign would naturally be grouped together. (This meant a different position than the e-key)

There was a long discussion on the this. It was pointed out by Knud Nordby from ETSI that this was a time when a standardized input of the Euro Sign on key pads could be done. Mike Ksar said this was important, but he wanted to get it verified that the Ermes code tables used by ETSI referenced the OCR-B standard. He wanted to know from Karl-Ivar why he did not like the way ETSI referred to the 10646 standard (Karl-Ivar had mentioned this earlier) and he said it was wrong to standardise before the sleeping industry decides anything. Erkki reminded Mike Ksar that ETSI had a strong industry participation (including telecoms).

There was a discussion on the Ermes/GSM code tables of ETSI. Evangelos Melegrakis said he would like to send the message to the ETSI community that the whole coding community saw the ERMES/GSM code table as a very strange animal.

Karl clarified, his position about how ETSI refers to 10646. He said this was only a technicality, he did not object to the content, just the wording. Graham Dixon asked about the input method of telephone key pads and why his telephone did it differently than the one Knud Nordby had. Knud explained that the standards were not mandatory and different companies were free to develop their own solutions.

There was a discussion on what the Euro Workshop could recommend and there seemed to be a consensus that the meeting should not state a specific position without consulting users and industry. Chris Makemson and Kenneth Thompson suggested seeking input from ANEC (European Association for the C-ordination of Consumer Representation in Standardization) and from ETSI itself. One final, sure resolution would be to suggest to put the Euro Sign on telephone key pads.

Karl - Ivar commented that the new edition of 10646 with Euro Sign in it was probably not known by ETSI and information on it should be forwarded. Erkki noted that on Monday 2 Feb a working relationship had been established through CEN/ISSS between ETSI and CEN/TC304 that could be used to forward such information.

Bar Codes

A discussion took place on the need to make the relevant technical committees in bar coding aware of the Euro Sign. It was proposed that this should be addressed. The secretary said that this was not needed and the relevant CEN committee was already well informed. Mike Ksar said this was not the case in ISO It was suggested that information be fed to ISO SC 31.

Fonts

Erkki said the question was if CEN/TC304 wanted to keep on our involvement with fonts. (whether the issue was closed to everyones satisfaction).

Karl-Ivar talked about the restrictions that may come from the EC on the shape of the Euro glyph. Erkki commented, that if the Commission had given up its status as a logo, it should be treated as any character would, and it was for the font industry to determine its shape in different fonts.

Mike Ksar said that this had been discussed before (Crete/Dublin), and that we were to differentiate between the logo and the type. The font industry should have freedom, he said that glyph and font issues were a done deal.

Lazaros Tossounidis commented on this. He said there was a logo that had been defined. The initial purpose was to create a logo. At some time in the process, - we have intervened and said OK there are a lot of things to be done in IT. (Originally it was very restricted use the Euro Sign had). In Dublin (october 97) we said we should have professional advice, and I understood variations could be specified. - Now we get comments, people see the glyph and do not like what they see. Users do not like the design by Monotype (for Microsoft).

Mike Ksar said it was important to note in the minutes that users in the Commission have not been happy with what has been produced. This should alert others.

There was a discussion on whether the Euro Sign was a capital or a small character. Karl-Ivar wanted it stated clearly that the Euro Sign was a capital. Otherwise it might get treated as a small c.

Lazaros Tossounidis said to this, that the document from the Commission on the usage of the Euro Sign made it clear that the Euro Sign was intended to be used in the same way as the Dollar Sign. He therefore said the Euro Sign is similar to the Dollar Sign and the Pound Sign in that there was no particular designs for small and capitals.

Lazaros Tossounidis asked for input from the industry experts present on when the Euro Sign would be available in printers. The answer HP (Mike Ksar) gave, was that system fonts could by used by Microsoft systems for printing. Lazoros Tossounidis explained and said that many printers had fonts in ROMs (that were not downloaded each time a printing took place). Lazaros Tossounidis wanted a date for when new ROMs would become available.

Tom Garland from Sun Microsystems said that Sun downloaded fonts the same way as HP, and was waiting for the printing manufacturers to provide ROMs.

Karl-Ivar said in Windows the coding for printing was deeply embedded in the system. He said we could not be too ambitious about getting printers that could print new code positions. Evangelos Melagrakis said that there was nothing in the operating systems themselves that forbid printing the 8-9 columns of the 8 bit code table (the column where Microsoft has put the Euro Sign). He said there might be filters and some obstacles that must be removed but there was no fundamental difficulty.

Karl Ivar came with information about Windows 3.2

Mike Ksar said that there was the solution for printers already existing and Lazaros Tossounidis agreed that this was a good transition solution. Evangelos repeated that in the core of operating systems there was nothing preventing the use of column 8 and column 9.

Tom Garland asked questions on the Edifact standard that only refers to the EUR and not to the Euro Sign( UN/Edifact is a standard for electronic data interchange for administration, commerce and trade, somwhat similar to ANSI X.12).

Van Wingen said that even that Edifact was not a character set standard, it was treated as such in the Netherlands were Edifact levels were referred to, in place of character sets. Erkki played down the importance of Edifact and said for his personal hat that Edifact was not a leading character set standard, in fact, not at all.

Finally on the input/output discussion on the Euro Sign. Mike Ksar recommended that TC304 contact the Braille person that came to Crete (from Austria) to see if the Euro Sign could be incorporated in the Braille system (for which the patterns are in 10646) This would be one of the tasks of the PT on the Euro.

7 Electronic code for information interchange:

16-bits. (10646 and its various subsets, such as MES/EES), UNICODE

It was noted that the proposal for inclusion of the Euro Sign into 10646 (Unicode) had met no objections and was being rushed through. There seem to be no objections to including the Euro Sign into European subsets of 10646. There was therefore no discussion on this in the meeting.

8-bit coded character sets - Preparation for the disposition of NB comments on the SC2 proposal on 8859-15 (³Latin-0²).

Erkki reviewed the status of the proposal for the 8859-15 8 bit character set (new code table with the Euro Sign, French characters and Finnish characters). The comments received from NBs, (to be discussed at a meeting in Seattle in March) had been split, some on the Sign and others were on the Finnish Characters S and Z with caron.

John Van Wingen said that it was a mistake to put three issues into one proposal. He said such proposals got voted down too easily. Mike Ksar wanted it noted that the secretary of SC2 had said (in email on the SC2 lits) that it was not true that Latin 15 had passed (in reply to messages that said it had). Keld Simonsen noted that at CD level there was not a precise percentage that decided if a proposal had passed or not.

Erkki said he change his hat (from chair to a Finnish NB representative) and he must state in reply to comments that it apears that the Finns had previously not asked for S and Z caron in Latin-1 that supposedly covers the needs of the Finnish language, that his predecessor apparently had not consulted the appropriate authorities in Finland, nor did he represent the current view of the Finnish National Body - I cannot undo what has happened, I can only say that once you know that errors have been made they should be corrected. Erkki then pointed to the paper from Klaas Ruppel on the S and Z caron that had been distributed among the participants.

Van Wingen said he must object to this subject being brought up in this meeting. He said it had not been advertised in any agenda that this meeting would discuss the subject of the Finnish characters. Marc Küster said that he wanted to have minuted once more the position of DIN, that it was not only sceptical about the need for the Euro Sign, but also for any new 8 bit code table, be it for the Euro, be it for the Finns, be it for the French characters. - The way to do it is in 10646.

Lazaros Tossounidis said he was mandated as user from the EC to declare a very strong interest in the 8 bit work that was going on. - It is obvious that for a couple of years we have to support the Euro in the 8 bit world.

Keld Simonsens said there was need for more language support and we only had 8 bits.

Mike Ksar said that it was extremely important to understand the severe impact the modifying of 8 bit code tables would have. - There are serious objections, the cost will be immense and it will end on the customers. Thus instead of creating additional parts or modifying existing ones, the solution that Microsoft has come up with to utilise the 8-9 columns even if it is not an ISO standard is a better solution. Most systems use this solution (as MS has the biggest part of the market).

Mark Küster wanted to send the following comment to the Commission: Even if Latin 0 succeeds, the old systems would not be able to use the Euro. The position of DIN was verified by a meeting that it had with industry which was in support of what DIN has said.

Van Wingen objected to the discussion of the columns 8 and 9 in ISO 8 bit code tables. He said he had put forward a proposal for this many years ago that had been voted down. He was not happy with hearing rumours about such a move now.

Evangelos Melagrakis (convener of ISO SC2/WG3 -8 bit coding) said he was behind the rumours of adding the two columns. He said he would put forward a proposal for this, but he had not wanted to interfere while Latin 0 was under voting. - It is true that this is an old idea that was voted down, but is has become relevant again. The proposal is to use the C1 for extra characters. We think it is time to come to vendors and give them opportunity to become compatible. Evangelos said this would make it possible for Telematics standard to become compatible as well as Windows code tables. He said this would be tabled in the Seattle meetings in March and that the proposal would be tabled prior to this meeting.

Karl-Ivar said that this proposals was not good enough, we needed to have a solution for 192 character systems. The EBCDIC code tables of IBM are already using the C1 slots in the code tables. Evangelos said that we should not be unhappy with solving 90% of the problem.

Mike Ksar said in response to Van Wingen about this proposal having been voted down before. He said the world had changed and that people that had been against the proposal before were now for it. Mike Ksar said that it would mean a new locale for UNIX system. For EBCID point of view, he shares the concern of IBM, but wanted to know what IBM would do if someone wanted to replace its codetables..

Monica wanted to add that IBM were in the banking and behind Windows systems and its presence there was not insignificant.

Borka objected to all this talk of 8 bit code tables and said that this committee (CEN/TC 304) had for many years promoted the use of 10646/Unicode, so why not use this opportunity and push for Unicode? Erkki gave an answer to this and said that standardisers should offer choices.

Chris said there was a huge amount of legacy system and then he introduced a person from Tandem (Compaq). He offered an action on himself of providing a statement of requirement for 8 bit character sets.

Karl-Ivar said we needed to concentrate our efforts to things that needed attention. (He was referring to his proposal of replacing the cedilla with the Euro Sign). Karl Ivar said IBM would replace a character with the Euro Sign (confirmed by Monica Ståhl). His proposal was for adding the Euro Sign in a number of 8 bit code tables, through replacement of the Cedilla, not necessarily by a new ISO standard but by registering such code tables or by making recommendations for their use.

The discussion was already getting very long and Keld Simonsen proposed we tried to close it. Lazaros Tossounidis said he agreed with Keld that we could not resolve all 8 bit problems in this meeting. Nevertheless he said we could have an agreement on that we wanted an 8 bit solutions. He added that we needed to act fast, and said there was a competition for resources because of the year 2000 problem.

Erkki said that he wanted to know if there were strong reactions against the 8 bit need for Euro solutions. Mike Ksar suggested that a list of alternatives was provided. He said it was forbidden to exchange one character for another in ISO code tables. Karl-Ivar said the problem was with the 192 character sets. Karl-Ivar wanted TC304 to state that it saw this as a problem. Küster reiterated the position of DIN. Monica said she agreed with Karl-Ivar.

Wolf said that CEN did not need to be slave of ISO and we could make a European resolution on using the C1 characters. Johan Van Wingen wanted to thank for the compliment he was receiving by putting his 10 year old solution back on the table.

Erkki wrapped up the discussion by saying that the 8 bit worlds needed solutions for the introduction of the Euro Sign. There followed a discussion on whether there were one or many solutions needed and Karl Ivar said that there was more than one problem with more than one solution.

The discussion on 8 bits ended without conclusions, but proposed resolutions were drawn up later the same evening.

7 bits

Graham Dixon raised the issue of the Euro Sign in the 7 bit world. He said the seven bit standards allowed composite characters and we could introduce the Euro Sign into them by a - c plus a backspace and an equal sign. Some of the delegates objected to entering 7 bit discussion and said 7 bits standards were dead and irrelevant.

Interoperability

Erkki said that interoperability had been on the agenda all the time because by introducing a new code table we created an interoperability problem. He said we did not need to address this any more specifically. Mike Ksar suggested a stronger statement, something like that mapping tables, filters etc. were solved by industry and that we should realise that there were problems with solutions outside the scope of this committee. Karl-Ivar strongly encouraged a further discussion on email lists. Keld Simonsen added that the Internet community should take part in the discussion.

8 Other aspects:

     Sorting, matching.

Keld Simonsen said we need not discuss the sorting of the Euro Sign because it needed to enter character sets before it got sorted. Þorgeir said this was not so, it was already in the sorting standard ISO 12991 that was soon to become a standard. He proposed this issue was handed over to the newly established Project Team on sorting and that the team to be set up on Sorting and Matching characters should look at mathching of the Euro sign in search engines, it was already being used in place of an e, in names such as Infeuro.

Writing of amounts

Keld Simonsen introduced his paper on writing monetary amounts in the 14562 draft (cultural registry), that addresses simultaneous rendering of amounts. Keld said that SC22/WG20 had done work on this and would work further on it. Keld said the Euro offered a window of opportunity for such work and a change for promotion and leadership.

Tom Garland said he had read the paper from Keld and thought it was excellent. - One thing I am not sure about, is the time dependency, (with the Euro introduction) as there is a need for dual currencies that does not have to be tied to the Euro, but can be used by any country having dual currency situation.

Keld wanted APIs to be tied with the cultural registries and to have comments from industry, users and supplies. -

Erkki thanked Keld for his contribution and said it was one example of how to address the problem.- We should have one statement about the general situation, dual currencies will be required. I take an example in Finland. If the base price is in Finmark, then the converted price in Euroe will be given with three decimals, hwereas a base price in Euro will need two decimals only. Wolf Arfvidson objected to this and said he had studied all the papers and said there was no requirement to give three decimals when going from one currency to another. Erkki said this obviously needed to be checked.

Graham Dixon said that there ought be an agreement on how to present the Euro currency and Erkki suggested this was work for the Project Team on the European Default locale. Others objected to this and said amounts would be written according to the current practice in each country.

Lazaros Tossounidis said Graham Dixon had a point and said travel between countries would increase, by people having the same currency in their hands and expecting the same rules everywhere.

There followed a long discussion by Erkki and Keld on API standards. Tom Garland said we needed to talk to users and get their input before writing the APIs. Graham and Keld talked about flexibility in the APIs. Tom Garland said that the flexibility could be built in at the application level.

Chris Makemson talked of the retail sector. He took an example of fuel pumps and said companies would not invest much in temporary systems. He said we expect one pump with Euros and other pumps using older currencies.

Tom Garland suggested introducing the Euro Sign into the Java locales. Erkki said that if we took up Java at this stage, we would have to take up many other platforms as well. There were no conclusion drawn from this discussion.

9 Reports on national Euro projects

Chris Makemson gave a report on UK and the Euro with emphasis on the market need for guidance. There was not time to review the Swedish national work.

Erkki Kolehmainen thanked the participants and closed the Euro Workshop meeting with an announcement of, and opening of another open meeting of the Euro Project Team where resolutions proposed to the TC304 plenary would be drawn up as the conclusions of this work.

10 Conclusions, recomendations

Resolutions:

The resolutions that were drawn up as an output of the Euro Workshop were the following (no 15-18 in the TC plenary, where all voted for them except Germany that abstained on resolutions 15-17 and voted against resolution 18).

15   CEN/TC304 recognises that there is no need to develop a European keyboard layout standard based on the Euro requirements only. However CEN/TC 304 recommends the implementation of the proposed recommendation from the EC commission (as described in doc N774) which matches what the IT industry and customers have requested. The recommendation suggests using the AltGr + e in most national keyboard layouts as the preferred solution, and CEN/TC 304 encourages its incorporation into the future editions of the ISO/IEC 9995 standard.
- Action on secretary to send this to ISO/IEC JTC1/WG5.

16   CEN/TC304 recommends adding the Euro sign into to the glyph repertoire of OCR-B (ISO 1073 part 2). This should be considered at the next meeting of ISO/IEC JTC1 SC2/WG3 in Seattle, March 1998.
- Action on secretary to forward this resolution.

17   CEN/TC304 expresses the need for the Euro Sign to be entered using telephone keypads. CEN/TC304 also recommends that ANEC (European Association for the Co-ordination of Consumer Representation in Standardization) and ETSI User Group provide a recommended best position based on customer and industry input.

18   CEN/TC304 recognises the need to provide the encoding for the Euro Sign in the 8 bit coded character set standards. CEN/TC304 urges ISO/IEC JTC1 SC2/WG3 to provide timely solutions.