Minutes of the WG10.5 meeting
Munich Friday 12th March, Saturday 13th March 1999
Recorded by Adam Pawlak

The meeting spanned over two days, and started at 17.00 in Room 106 of the ICM - Convention Centre in Munich on Friday, 12th of March. It was continued on the following day at the Technical University of Munich. The meeting was chaired by Jean Mermet.

All group members are very grateful to Pr Kurt Antreich for his kind hospitality, in his department at the University. He offered perfect conditions, nice room, drinks, snacks, etc.., for a pleasant and fruitful brainstorming of the group. He took care of everything by himself, as no personal was on duty on Saturday.

We also direct warm thanks to Adam Pawlak who kindly accepted to keep track in writing of all the discussions over the 2 days. These minutes are due to him.

W.G. Meeting, Friday 12th of March, 17.00-20.00, ICM, Munich

Attendance:

The meeting on Friday was attended by: Dominique Borrione, Klaus Buchenrieder, Edward Cerny, Carlos Delgado-Kloos, Hans Diener, Nikil Dutt, Martin Edwards, Hans Eveking, Norbert Fristacky, Manfred Glesner, Werner Grass, Reiner Hartenstein, Ahmed Jerraya, Thomas Kropf, Odysseas Koufopavlou, Jean Mermet, Klaus Mueller-Glaser, Wolfgang Nebel, Adam Pawlak, Robert Piloty, Franz Rammig, Ricardo Reis, Michel Robert, Donatella Sciuto, J. Staunstrup, Alexander Stempkovsky.

Item 1 - Approval of agenda

CODES workshop, and IFIP sponsorship of DATE were added to the agenda.

Jean Mermet announced a discussion on SIGs on Saturday.

Item 2 - Approval of minutes of Palo Alto meeting

Minutes of the previous meeting in Palo Alto were accepted.

Jean Mermet announced a message from Ganesh, who wanted the "in co-operation status" for Async'99, scientific topics yet to be proposed by him. Jean Mermet thanked Steve Johnson for writing the minutes, and Nik Dutt for organising and chairing the W.G. meeting in Palo Alto.

Item 3 - Self introduction of O. Koufopavlou and M. Robert

Jean Mermet recalls participants of the procedure decided for extending the W.G. with new members. First, applicants must send their candidacies in writing. Second, the group examine the candidacies, with help of 2 reviewers. Then candidates voted as eligible are invited to introduce themselves in person during the next W.G. meeting.

The candidacy of Eugenio Villar has been received positively already in Palo Alto, but, due to sickness, Pr. Villar could not travel to Munich. He will be invited at either the DAC meeting or the following meeting in Portugal.

Dr. Koufopavlou and Prof. Robert present at the meeting were asked to introduce themselves.

Dr. Koufopavlou introduces his work, saying among others, his main research area was "low power design". Later, Prof. Michel Robert introduced his work. (See attached slides).

Item 4 - Final discussion about these candidatures

While the two candidates left the room, Jean Mermet encourages a discussion. He informs that Dr. Koufopavlou was recommended by Prof. Goutis who will resign if the candidate is accepted.

Voting for Dr. Koufopavlou brings 0 against his candidature, and 1 abstain.

Voting for Prof. Robert brings 0 against, and 0 abstain.

O. Koufopavlou and M. Robert are then invited to join the meeting as regular members.

Item 5 - Report on SIG-VHDL

Jean Mermet informs that the chairman of SIG-VHDL, Wolfgang Nebel, has the intention to resign soon. Wolfgang Nebel SIG chair then reports on the activity of his group.

FDL'98 in Lausanne was a success due to the action of Alain Vachoux, chair of the event. It was mainly a success from the scientific point of view, not from the financial point of view. Wolfgang Nebel would consider it as an investment. He proposes to vote on the SIG-VHDL report later on since there will be a new related item in agenda.

On behalf of 10.5, Jean Mermet expresses his appreciation to Wolfgang Nebel for his chairing of the SIG-VHDL and to Alain Vachoux for his excellent job to launch FDL.

Item 6 Report on SIG-CODES

Martin Edwards is the new elected chair of SIG-CODES.

Klaus Buchenrieder, past SIG-CODES chair, thanks all SIG-CODES members. According to him, SIG-CODES has integrated the co-design community. The CODES/CASH workshop belongs to the SIG and any event with this name must be sponsored by IFIP 10.5. Since the workshop has grown substantially and gathers over 100 participants he suggested to think of it as a possible "working conference". If the organisers don't want the IFIP sponsorship anymore they will have to change the name.

On behalf of 10.5, Jean Mermet expresses his appreciation of Klaus Buchenrieder's work as chairman of the SIG-CODES.

Ahmed Jerraya, general co-chair, explains the state-of-the-art in planning for CODES workshop in 1999. They are late with the IEEE application forms. Klaus Buchenrieder suggests to accept "new forms" for this workshop. He explains also the SIG-CODES Steering Committee membership rules.

Jean Mermet explains the procedure for IFIP sponsorship of events. The form should go first to the 10.5 group chair, then to the TC10 chair, then to the IFIP secretariat. This procedure can be quite quick if necessary. Jean questions Ahmed for reasons of withdrawing application for IFIP sponsorship. Ahmed blames IFIP bureaucracy. Jean explains that IFIP procedures is in fact very simple but probably not well explained in the past. Ahmed replies that SIG-CODES will reconsider application for IFIP sponsorship.

Jean Mermet puts a motion : "Is WG10.5 in favour of continuing to be the main sponsor of the CODES workshop ?". The result of the vote is: 0 against, 0 abstain

The W.G. has voted for sponsoring the CODES, this year without financial obligations, next years possibly with financial obligation. The final program should reflect this status.

Item 8 - Report on VLSI99

Luis Miguel Silveira from INESC, presents the preparations of VLSI99 that he volunteers to chair and organise.

Although the location of the event is Lisbon, the final decision on the place of the conference has not been met yet. Three options: Technical University of Lisbon, or one of two hotels. The decision will be taken by the end of March 1999.

The structure of the event: 3 days event, two tracks, tutorials at the last day of the event.

The Technical PC is presently being built. Reviewing will be done in electronic form.

As it is the rule for any 10.5 event that any member of the group can be a member of the Program Committee, Jean Mermet suggests to ask by e-mail WG members to join the PC of VLSI'99.

Interlude - K. Mueller-Glaser's proposal to report on a possible "SIG - Microsystems"

The change in agenda was accepted to allow Prof. Mueller-Glaser, who had to leave, to address the W.G. Mr Mueller-Glaser, chair of IFIP W.G. 10.7 "Microsystems" explained shortly the aim of the W.G. 10.7. Apparently, 10.7 is not dedicated to technology. CAD for microsystems is not that much different from what he thought 10.5 was covering in its scope. He suggested to have a new SIG " microsystems" within the WG 10.5 in replacement of 10.7.

Jean Mermet proposed to put this item on the agenda of the Saturday continuation of the meeting.

Item 9 Report on FDL99 by Jean Mermet

FDL99 will be organised by ECSI and the Ecole Normale Supérieure de Lyon. It is one goal to have FDL99 cheap for attendees. Jean Mermet was looking for a good location and cheap organisation. He was looking for sponsors. The budget is higher than for FDL98 (printing and mailing of the "call" and preliminary program will cost aprox. 30K Euros). MEDEA will mail the preliminary program, Mentor, Synopsys, Cadence will mail 15000 programs. The program is planned to be distributed to the DAC participants in their bags. These programs could be printed in the US.

Cheap participation of students: low registration fee (half the regular fee) and cheap accommodation in students dormitories, have been set to allow more student participation.

Item 10 Report on APChDL99

Matt Imai was not present. Jean Mermet had no information on APChDL99., but statistics on APChDL98 ( 85 participants, among them 70 from Korea, 1 Europe, 1 US, ...). Increasing international participation is definitely an issue

APChDL2000 will be in Beijing as one track of the IFIP World Congress.

Item 11 CHARME99

The CHARME workshop will take place on Sept. 27-29 1999 in the Black-Forest, in Germany.

Thomas Kropf, general chair of the event, informs that the workshop is a fully sponsored IFIP event. Several industrial co-sponsors (e.g Intel, ) support the event.

35 submissions were received 3 days after the deadline. 15-20 paper are to be selected.

Springer-Verlag Lecture Notes will publish the proceedings of the workshop. This is reported to be in contradiction with IFIP rules that requires publishing at Kluwer.

Jean Mermet wants to definitely correct this erroneous interpretation: IFIP wishes preference to be given to KLUWER only if they offer conditions at least as good as others. If there is a clear advantage to sign with another publisher the editors should not hesitate.

The meeting was adjourned at 20.00.

W.G. Meeting, Saturday 13th of March, 9.00-15.30, Technical Univ. of Munich

Attendance:

On Saturday the W.G. meeting was continued at the premises of Prof. Kurt Antreich at the Technical University of Munich. This meeting was attended by: Kurt Antreich, Dominique Borrione, Eduard Cerny, Carlos Delgado-Kloos, Hans Diener, Nikil Dutt, Martyn Edwards, Hans Eveking, Norbert Fristacky, Manfred Glesner, O. Koufopavlou, Thomas Kropf, Wolfgang Nebel, Jean Mermet, Adam Pawlak, Franz Rammig, Ricardo Reis, Michel Robert, J. Staunstrup

Opening the meeting Jean Mermet stressed that not only the organisation of events, but also some projects with scientific activities could be pursuit by the W.G. 10.5. He referred to the CONLAN project as an example. New activities could be run in new SIGs. He advocates for establishing new SIGs, the W.G. should define general policy, SIGs and projects do the scientific work.. Further, he postulated videoconferences 2 or 3 times a year.

Item 12 DATE98, DATE99 by Jean Mermet

The WG.G 10.5 is a sponsor of DATE. This was decided several times by the group but never achieved. Jean Mermet said that 50% of sponsors didn't bring any money, only two sponsors did: the EDAA Association and ECSI. Other sponsors brought guaranty against losses.

He required from DATE organisers to fill in the IFIP forms . The process is on its way.

DATE earned 300K Euros. DATE will keep 200K$ in cash for next event organisation.

ECSI gave a grant of 12K$. The WG 10.5 will follow its regular way and give a loan of 6K$.

Manfred Glesner said that DATE was very expensive. He claimed lower prices for universities. According to Jean, it was also expensive for exhibitors. A recommendation should be passed to the DATE organisers to lower the prices for universities. Dominique Borrione postulated that authors also had lower registration fees. This is seconded by Wolfgang Nebel. The DATE organisers didn't care for lower hotel prices either.

Franz Rammig recommended to influence the DATE Steering Committee through its IFIP members.

Point of order

On the proposal of Dominique Borrione the order has been changed. The future of SIGs is to be discussed at first.

Item 19 Re-shaping the activity around SIGs

Martin Edwards, the new chair of SIG-CODES explained his new strategy for the SIG-CODES.

SIG-CODES was established in 1994. To sponsor the Int. Workshop on HW/SW co-design, transformed later into CODES/CASH workshop.

SIG-CODES has 20 members since the beginning. No new members, nobody leaving.

Martin wants to increase the membership of the SIG by inviting new members to join.

The rules of SIG-CODES are OK, no need for changing. The profile of the SIG is well identified, and is not copied elsewhere. Big conferences have sections on co-design.

There is no reasonable Web-site. He intends to build a new Web site with reasonable explanation of what co-design is.

Nikil Dutt questions about the role of SIGs within 10.5. Jean Mermet tells, he has no answer.

Further, he explains that SIGs were originally promoted by Egon Hoerbst to organise events: SIG-VHDL around VHDL, SIG-CODES for co-design. Wolfgang Nebel adds, there is no need for being members of the W.G to be member of the SIG which makes them more flexible structures, in principle open to young researchers..

Nikil Dutt considers SIGs as "discussion groups" in addition to event sponsors.

Jean Mermet has asked whether other W.G. members wish to participate in SIG-CODES ?

For Nik Dutt, the CODES Workshop is no longer a workshop, rather a Working Conference with a more general scope.

Further, Wolfgang Nebel reported on SIG-VHDL. It originally recruited 30-40 members, later remain very stable. The VHDL-Forum, earlier the main event of SIG has merged into FDL. The SIG-VHDL is now open to Verilog and other HDLs. SIG-VHDL wants to extend its scope to SIG-HDL or SIG "Design Languages". New people have to be motivated to take over responsibilities.

Jean Mermet suggests to start "pilot projects". A SIG can organise a pilot project, apply to EC for support (workshops, collaboration). Wolfgang Nebel questions such an action for VHDL, perhaps on HDLs... Wolfgang points to a WG on "low power".

The European Commission can help projects. Six members received money from EC for workshops and events.

Jean Mermet continues on "pilot projects". He points to VSI Alliance pilot projects on testing the new standards in a real industrial environment. New domains in research, new domains for the application of the standards are open.

Jean Mermet observes that international collaboration doesn't start from scratch, collaboration starts after many meetings. Someone has to take the initiative. The SIGs could be open to partners from Central and East European countries, to initiate collaboration.

Thomas Kropf : "SIGs should have more industrial collaborators".

Eduard Cerny: "industrial partners don't show up".

Thomas Kropf underlines that in the US there is more industrial participation in SIGs. Norbert Fristacky stresses the difficulty of introducing industry to SIGs.

Jean Mermet adds that although W.G. has smaller industrial participation, SIGs should have more industrial participation. SIGs could be co-chaired by industrial and academic partners. Bylaws of SIGs should be simple, and flexible, with simple procedures for joining and leaving the SIG.

Franz Rammig complains that there is no right location for "hot topics" in IFIP, ( e.g. on "embedded systems"). He suggests to approach J.C.Laprie, chair of TC 10 to form an appropriate SIG covering "Embedded systems". SIGs could be co-sponsored by several WGs.

Jean Mermet seconds the idea: " interesting topics are at boundaries".

For Juergen Staunstrup, the activities of SIGs should make possible the involvement of post docs, more than of "people traveling already a lot".

Thomas Kropf -SIGs should be open to new members, and have a clear procedure to finish membership of non active members.

To improve interest and organisation, which may result in a better attendance, Jean Mermet suggests that at conferences sponsored by IFIP W.G, as a new general rule, one day should be planned for meetings of SIGs. Organisers should take care of rooms and facilities.

In addition of the 2 existing SIGs of 10.5, the SIG-VHDL and the SIG-CODES, new SIGs are proposed:

SIG "Embedded systems"

Franz Rammig expressed an interest in a potential SIG on "Embedded systems". Such a SIG could be created jointly with other WG, perhaps even with an IEEE group?

SIG-CHARME

Dominique Borrione explained the history of CHARME which started in 1984 with a workshop on formal methods applied to design, organised by Hans Eveking and George Milne. CHARME was the original acronym of an ESPRIT project. Later appeared HOL - Higher Order Logic Workshop, DCC - Designing Correct Circuits Workshop, all in Europe. In the US was created FM CAD in Silicon Valley ('96, 98). As a result of an informal agreement, workshops on formal methods are organised biannually, once in Europe once in the US.

Dominique Borrione observed that the FM CAD Conference is on industry oriented aspects for formal methods in CAD, well supported by the US industry.

Ed Cerny thought that US members might be interested in having SIG-CHARME, although they have not yet acknowledged the proposal..

Dominique believes that the formal methods community would like to merge DCC, CHARME workshops into one, single event on format verification in Europe. Thomas Kropf agrees .

Hans Eveking seconds the idea of SIG-CHARME.

Dominique Borrione reminds of Steve Johnson's action devoted to education on formal methods. He might be interested in becoming a member of SIG-CHARME to develop his initiative.

Jean Mermet summarises the potential activities of SIG-CHARME:

- education on formal methods

- sponsoring an event every two years (alternately with FM CAD),

- one full day for meetings,

- brainstorming on new standards.

Dominique Borrione will contact Merry Sheeran about the DCC workshop.

Nikil Dutt re-formulated the goals for SIG:

- open to Ph.D. students and industry people

- mission to work technically,

- events are not of primary concern and should include technical working meetings.

Jean Mermet welcomes this formulation of this SIG mission which concentrate on technical issues. SIG should be a place where young people can join.

To start the process of industrial co-chairing, Jean asks Thomas whether he would co-chair the SIG-CHARME, Thomas Kropf confirms his interest, but he needs preliminary OK from Bosch.

Dominique proposes to contact Warren Hunt and Steve Johnson to invite them to the SIG. During the next W.G. meeting at DAC Dominique, (who is not willing to chair the SIG) will be commissioned to find the other co-chair of SIG-CHARME.

Action item:

Dominique Borrione and Ed Cerny will prepare a position paper on SIG-CHARME.

SIG on "Embedded systems"

For Franz Rammig, a SIG on "Embedded systems" should be more oriented towards: SW, RT operating systems, and code generation, which are more important than hardware (industry takes standard micro-controllers, and concentrates on SW). It must include people from control theory, SW community plus operating systems. Definitely, there is a need for a bridging activity. These are topics across different IFIP WGs.

Action item

Franz Rammig will prepare a text on "Aims of SIG - Embedded systems".

Later, the W.G. will decide who contacts TC chairs.

Wolfgang Nebel expressed his concern about a possible overlap between "Embedded systems" and HW/SW co-design. Franz Rammig responded that embedded systems were application driven.

Nik Dutt added, the scope of SIG-CODES was very narrow.

To prevent ambiguity, Jean Mermet suggested a new name for the SIG - Application Driven Embedded System (SIG - ADES).

SIG-" Microsystems"

Jean Mermet referred to the proposal of Mueller-Glaser, on Friday, questioning whether W.G10.5 needs a SIG devoted to "Microsystems" (sensors, actuators on a chip). If the answer is "yes" there is a need for industrial participation in this SIG.

Manfred Glesner explained activities on micro-systems in Germany. He leads the GI (Gesselschaft fuer Informatik) activity on micro-systems. He confirmed that the area was under-developed. Hans Diener added, in this area one needed a national project. Manfred Glesner further observed, that nobody exploited it from the system level.

According to Jean Mermet, there exists a W.G. 5.2 on mechanical CAD which could be interested.

Franz Rammig didn't see the reasons why a not growing W.G. 10.7 should develop better as a SIG of W.G. 10.5. Why this change should help this group ?

It seems there were no serious activities in this group. Wolfgang Nebel added, that the group was not technology driven.

In summary, it was not clear why change to SIG of W.G. 10.5 could help W.G. 10.7

Manfred Glesner suggested to organise at VLSI99 one session on "Microsystems", and Euro-practice-like activities for "Microsystems".

Action item

Manfred Glasner will ask Mueller-Glaser for more explanations.

Study group on "System specification issues"

Finally, Jean Mermet expressed his interest in creating a study group dealing with "System specification issues". He explained the goals of the SLDL workshop. SLDL as a W.G. aims at producing a set of system level notations. Jean felt they lost momentum. Firstly, they were motivated by EDA Roadmap Industry Council, now more motivated by topics like "connection driven synthesis". He would be happy to give an umbrella to this activity.

The SLDL subcommittee "Models of computation" didn't start any real activity. Jean would like to study models of computation, a good starting point to studying system notations. He is interested in using "B" for this.

Jean has shown transparencies explaining the basic concepts of "B" (based on set theory) Apparently, STERIA, a French company has built tools, that are well supported. They were used by MATRA in Paris to create the software of METEOR, the "metro without driver". In B, a designer is looping a lot at the specification level. Further, Jean explained some basic ideas behind B specifications. B is also interesting in the context of system level IP reuse. Finally B could perhaps provide a bridging semantics for existing system level languages or notations.

Hans Eveking commented that the B approach seemed to be very ambitious.

Carlos Delgado Kloos thought that Jean had proposed a project rather than a SIG.

Dominique Borrione commented there was a need for a full time research project.

Franz Rammig thinks that the W.G. could provide a surrounding work to the proposed project.

Jean Mermet stressed the need for full time young people working on that topics. He concluded, that there could be kind of a "pilot project" reporting to SIG-CHARME, and to SIG-Embedded systems. He will try to set a project on this. Later, presentation will be made to both SIGs.

Item 14 - CHDL99

At CHDL99 - not enough papers were received. Ron Waxman proposed to cancel the event.

Franz Rammig thinks that the title of the conference is not attracting young people.

Jean Mermet points to FDL - HDL conf. in Europe, and to HDLCON an annual event in the US, more industry oriented.

Ricardo Reis thinks that the promotion of CHDL99 might be not adequate.

Jean Mermet questions whether there is a need for a CHDL conference moving around the world.

Dominique Borrione points to Asia Pacific DAC which was not successful, unless in Japan.

Dominique says that a good conference had to be well published (e.g. special issue of some journal). A good strategy is to organise something which is "soft" in requirements - wider contributions, and at the same time "tough" on selecting good papers to be published in a journal. Dominique proposed to extend the scope to accommodate new topics.

Franz Rammig proposes the concept of a "virtual conference". For CHDL we could try a new formula: a series of books.

Action item

Jean Mermet will talk to the sponsors of successful HDL conferences about a series of books regrouping the best papers from APChDL, HDLCON., and FDL.

This could be a means to perpetuate the tradition of the CHDL conferences and would be attractive to the authors..

FDL2000

Will be organised by FZI at Karlsruhe or Tuebingen.

APCHDL2000

Will be organised in conjunction with IFIP World Congress held in 2000 in Beijing. Although being simply a track in the IFIP world congress, APCHDL will keep its identity and PC.

VLSI2001

Ricardo Reis reminded that VLSI2001 was planned for May 2001 as an accompanying event to IEEE ISCAS Conference. Jean Mermet shares concerns about the organisation of VLSI 2001. He suggested, at least, to change the name of the VLSI Conference

According to Nik Dutt, in Lisbon a CALL for VLSI 2001 should be re-opened. If for VLSI2001 a group of dedicated people will not be found, then the conference should be cancelled

Dominique Borrione cited Richard Newton that the VLSI Conference was originally thought for papers with "crazy ideas in VLSI design", this is no longer the case, this is why the attendees of the Palo Alto meeting proposed to cancel it.

Adam Pawlak pointed to new ideas stemming from the opening of engineering environment to Internet. Collaborative and distributed engineering bring nowadays a high potential for new design methodologies, specially for system on chips the present avatar of VLSI.

Carlos Delgado Kloos supported this observation, also Jean Mermet seconded it.

Ricardo Reis thinks that VLSI combines important topics. He will discuss the issues of the VLSI Conference with the community. Jean Mermet suggests to create a advisory group formed from past PC chairs to collect opinions on the future of the event.

Nik Dutt suggests getting contact with previous PC chairs to come with a proposal at the IFIP meeting during DAC.

Nik Dutt suggests to contact Richard Newton and Petra Michel for the past VLSI conferences.

CHARME2001

Dominique Borrione, is not aware of any plans at date.

Jean Mermet enumerates the events owned by the W.G.: APChDL, CHARME, CODES/CASH, FDL, "new VLSI".

The rule proposed by Nik Dutt for all events:

- W.G. members are to be questioned whether they want to participate in the event's PC,

- W.G. members are expected to join at least one PC of their selection, and actively support

the event (participating in PC videoconferencing, but face-to-face meeting is also important).

Guides to PC chairs should be proposed: "how to organise a PC meeting". We have to study how to organise telecon- meetings, and find where, and how this is possible,( e.g. at IBM or HP) with sound and image. It was suggested that Francisco Corella could check with HP, Ed Cerny will contact him. John Darringer could be questioned about IBM. How about Siemens and Philips? Jean will inquire. Ed Cerny will check with Dave Agnew (Nortel).

Item: Location of meetings

The rule proposed, one meeting in Europe, US, Asia per year. Participation from Asia is very low. Ed Cerny suggests to increase Asian participation mainly through the SIGs.

Membership

In Palo Alto, a two levels participation was proposed. Jean Mermet doesn't like this idea. He doesn't want to accept "low level" of participation. He prefers that the group formulates clear rules. Non active members should not be kept in the group forever. In fact, it has been accepted last year that withdrawn members could easily become members again, after clearly stating their new motivation.

Jurgen Staunstrup suggested, for future meetings, to put in the minutes, the names of those members who didn't react, even by e-mail.

The W.G. members present at the meeting expressed their concern about some members of the W.G. who did participate in DATE and are in this situation.

According to Nik Dutt active members contribute to the IFIP mission. This could be recalled to non-reacting members. The chair will friendly contact them (by phone or letter).

SIG-Membership

Jean Mermet thinks that the procedure for the SIG membership should not be formalised to keep flexibility. The new chairs will set their own procedures and GO.

SBCCI99

Ricardo Reis asked for co-sponsorship of IFIP "without financial implication" for this event.

The voting brought: 0 Against, 0 Abstain.

Jean Mermet has explained that organisers may have a different publisher than the official publisher of IFIP which is Kluwer, providing the organisers have better deal than Kluwer offer from an other publisher.

Jean Mermet transmitted the message of Ganesh who is suggesting to keep a track of the W.G. members who join "big events". What about "white boards" to put messages of the W.G. members (Dominique Borrione) ? Nik suggested to include in the programme (of IFIP sponsored events) informal meetings of W.G. members. This could well be common lunch, or coffee.

Jean Mermet introduced the candidacy of Serafin Olcoz, coming from an SME called SIDSA.

He would like to join the W.G. Dominique Borrione points to his early work on VHDL formal verification. She seconded his membership. Carlos seconded the candidacy as well.

A vote was called: 0 Against, 1 Abstain. Serafin Olcoz will be invited to come and introduce himself at the next meeting.

Item 22: Next meeting at DAC

Jean questions the usefulness of 2 hours meeting which is the usual case at DAC. He doesn't want to continue with short - 2 hours meetings.

The following one should be the least one day long. It could take place during VLSI99 in Lisbon, in Portugal. A good idea will be to start the meeting on Friday evening and continue on Saturday as was done in Munich. Jean proposes not to have any meeting, either at DATE2000, or at DAC2000. These events are aimed at other businesses and WG members are too busy.

A rather small event in the US would be a good solution (proposals welcome).

A meeting during IFIP World Congress in Peking in 2000 is also planned.

The meeting was adjourned at 15.30.