Internet Architecture Board


IAB Minutes 1994-07-27

Home»Documents»Minutes»Minutes 1994»IAB Minutes 1994-07-27


    The open meeting of the Internet Architecture Board at the Toronto IETF opened with a report on the status of the liaison with SC6 by Christian Huitema. The “Proposed Cooperative Agreement” is almost complete, with only one issue still to be resolved. Christian then mentioned the request from the ISOC board of trustees to the IAB to devise a “code of ethics” for the Internet community.

External (Mis)Perceptions of IETF/IESG/IAB/ISOC

    Brian Carpenter presented a report on “External (Mis)Perceptions of IETF/IESG/IAB/ISOC.” He has been polling people inside and outside the Internet community to learn what the image is of these organizations. In summary, he has heard from some people that the IETF standards process is too informal, from others that it is too formal, and from a few that it is ineffective. The internationalism of the ISOC and IETF is viewed as inadequate. Some people view IETF standards as non-authoritative and the IETF/IESG/IAB as self-perpetuating with closed vision. The activities of the IAB and IESG are poorly understood both inside and outside the IETF. ISOC has over-emphasized political correctness and Internet hype, but has overlooked outreach to IETF and staff engineers. Regarding the issue of the internationalism of the IETF, it was suggested that there be more IETF meetings outside of the United States. It was not clear that non-US locations would be as cost effective for attendees, but it might address the perception that the IETF is primarily a US-centric organization. An Internet draft will be published shortly that details the comments received and outlines some suggestions for improvements.


    Next, Steve Crocker gave an update on the POISED process. A question raised during discussion was whether there is anything that is outside the purview of the IETF. Also, the statement was made that much of the focus on openness and process is less important than the technology–the reason that TCP/IP has proven so popular is that the technology works.

    There was then a prolonged discussion about the fact that some working groups are not working and that some people are choosing to not bring technology to the IETF because of certain aspects of the IETF culture. Possible causes for these problems are that people can’t be removed from working groups and members are not held accountable. One suggestion to fix this within the current structure is to have strong WG chairs and area directors. Some members of the IAB voiced the strong opinion that the IESG area directors should take stronger steps to manage the process. The area directors can and should take a strong role in managing working groups (e.g., by policing the charters and instructing working group chairs on their rights and responsibilities in running meetings).

    Another statement that triggered debate is that it’s important to have a clear vision of what the goal is and to understand how the work within the IETF is working toward reaching the goal. It was felt that the IAB should clearly define the architecture so that the IESG, IETF and the working groups can engineer the solutions to fulfill the vision.

Follow-on to the Security Workshop

    Christian Huitema then presented a talk “Follow-on to the Security Workshop.” He announced that the IAB, in its meeting on Sunday July 24, had developed a statement encouraging the IETF to develop security for the Internet:

      “The IAB calls on the IETF to develop standards that provide security (confidentiality, authentication and integrity) for communication over the Internet. The framework must make provision for multiple encryption algorithms, but should specify a single common (strong) method. Consideration of export controls is outside of the purview of the IETF.”

    The talk evoked a plea from the audience: the workshop report looked good and provided a nice perspective on a few topics, but the report is no substitute for the IAB providing an end-to-end top-to-bottom architecture for Internet security. This raised the question whether the PSRG is documenting such an architecture. There is a document, currently 178 pages long, but a lot is still in outline form. It will (soon?) be released as an Internet draft, but first it needs experts from the areas to help fill in the protocol-specific parts. There was strong encouragement from some quarters for near term release of the document to the larger community for discussion and input.

IAB Information Infrastructure Workshop

    John Romkey announced plans for the next IAB workshop, to focus on “Information Infrastructure.” There is a call for white papers, which appears below:

      IAB Information Infrastructure Workshop
      Call For White Papers

      The Internet Architecture Board is sponsoring an “Information Infrastructure” workshop Oct. 12-14, 1994 at an east coast site in the United States still to be determined.

      This workshop is intended to explore architectural issues involved in various information applications and systems currently being used on the Internet, including the World Wide Web, Gopher, Mosaic, archie and WAIS. The workshop will emphasize common architectural issues such as:

        type conversion
        security issues (privacy, authorization, authentication,…)
        resource location
        integration with directory services

      Participation in IAB-sponsored workshops is by invitation. Traditionally, attendees are IAB members, IESG area directors concerned with the subjects, and selected experts from IETF working groups and IRTF research groups. In order to solicit new perspectives in this area, the IAB invites interested parties to submit a one to two page white paper analyzing a specific architectural problem with architectural suggestions for solutions. Please email white papers to ii-white-papers@ELF.Com by August 31. They can also be sent as hardcopy to John Romkey, ELF Communications, 1770 Massachusetts Ave. #331, Cambridge, MA 02140, USA.

      Note that the number of places is very limited – there is no guarantee that we can invite all those who submit. But in any case, your comments will be taken into account during the workshop!

Attachment – Slides


        Open IAB Meeting
        Toronto, 27 July 1994


    • Progress of the liaison with SC6.
    • The ethics of the Internet.
    • The image of the IAB/IETF. Brian Carpenter.
    • The progress of POISED2. Steve Crocker.
    • After the retreat of the security.
    • Before the retreat of the applications. John Romkey.

    Liaison with ISO-IEC/JTC1/SC6

    • Requested by ISOC, 92.
    • Amsterdam 1993: Requested the writing of an MOU + discussions.
    • December 93 – January 94: A lot of discussions. MOU get drafted.
    • March 94: Presentation to the IETF.
    • May 94: Last Call.
    • June 94: Approved by SC6, completed.
    • July – August 94: Approved by IAB, Last Call. Submission to ISOC’s trustees.
    • November 94: Approval by JTC1.

    What changes in the MOU?

    • The title:
      Cooperative Agreement

    • The unspecified parts, “to be filled up by SC6/JTC1.”
      Selected groups send documents in electronic form to JTC6, then to point of contact.

    • A precision:
      Pending deliberations in JTC1, SC6 standards may include “normative” as well as “informative” references.

    The ethics of the Internet

    • Requests from the Internet Society’s trustees.
    • Devise “code of ethics”:
      • what users should do,
      • what providers should enforce.
    • This will be handled by the IESG.

    After the security retreat:

    • Start research work,
    • Start engineering work,
    • Pass the good word:
      • Protect SNMP, DNS, routing.
      • No password in the clear, never.

    But this is sad…

    • Recognize that we have no privacy,
    • Protect the password,
    • Leave everything else in the clear for hackers to prey on.

    …we should do better:

    • Specify “the right thing,”
    • Even if export controlled:
      • get international developments,
      • import instead of export,
      • get control removed.


    External (mis)perceptionsof IETF/IESG/IAB/ISOC

    • Some concern exists about this in the IAB

    • An informal survey of IAB and IESG membersand their personal contacts was conducted by email and phone

    • Now follows a sorted list of views expressedby various people; more views welcome

    • Anonymity respected; don’t blame thereporter for any of what follows!

    • Recommendations for any action TBD (will beposted as Internet Draft)

      July 1994 Brian E. Carpenter

    Perceptions of process (1)

    • IETF is a gang of hippies, too informal

    • IETF openness is a virtue

    • Openness allows anybody to filibuster

    • IETF rejects ideas without formal procedure

    • Lack of formal votes in WGs means thatpeople who are shouted down just give up

    • Process fails to produce well-engineeredstandards

    • IETF became bureaucratic and ineffective

    • Officials mistrust culture of competition

    Perceptions of process (2)

    • IAB & IESG seen as self-perpetuating

    • TCP/IP cheated on OSI by bypassing formalinternational process (trade policy view)

    • IETF should be open to outside submissions wanting rubber stamp

    • New WGs are formed bottom-up with gaps and overlaps

    • Unclear how comments on last calls are handled

    • On-line doc of process is good (but not forpeople who are not on-line)


    • IETF is not as international as ISO and ITU

    • IETF lacks formal international affiliation

    • Difficult for European officials to view ISOCas international (no EU finance)

    • Most government officials in [country X] haveno knowledge of IETF, little of ISOC

    • But some officials ask “how can we help?”


    • IETF is less authoritative than, say, the ATMForum

    • People outside think IETF is just like OSF orATM Forum

    • People think it is smaller & smarter than it is

    • IETF should not have monopoly on standards

    • People outside think IAB has power 🙂

    • People have not heard of IESG

    General perceptions

    • Closed perspectives – “IETF Gestalt” (e.g.ignoring economics, insisting on datagrams, knee-jerk reaction against OSI)

    • ISOC officers with industry jobs in conflict?

    • Speakers on Internet should avoid hype

    • ISOC is too PC, has ignored staff engineers

    • IAB has weak profile (since Kobe?)

    • IETF wonders what IAB does

    • IETF wonders why IESG is slow, insular andautocratic

    • IETF is too remote from users and operators


    POISED Process for Organization of Internet Standards

      Steve Crocker Trusted Information Systems

      Mel Pleasant Rutgers University


    • Selection of People

    • Standards

    • Charters

    3 Documents

    • 1601 IAB Charter

    • 1602 Internet Standards Process
      • 1602bis draft-ietf-standards-processv3-00.txt
    • 1603 IETF Working Group Guidelines


    • IETF Charter

    1640 POISED Report

    • Nominations Process
      • draft-ietf-poised-nomcomm-00.txt
    • Others as needed

    Process Changes

    • IESG became final arbiter of Standards

    • IAB became court of appeal for process issues

    • ISOC Trustees are final appeal

    Selection Changes

    • Two year terms

    • Nominations Committee
      • Non-voting Chair appointed by ISOC president

      • Seven voting members chosen at random from volunteers

      • Reps from IAB, IESG and IRTF

    Initial Experience

    • Two rounds of selection
      • Much turnover, as was needed

      • Very rough edges
    • Standards process is smoother but not perfect
      • More responsive

      • Increase in size has recreated delays

      • Semi-frequent end-runs

    Selection Issues

    • Non-voting chair and reps

    • Chair’s role w.r.t. inactive/uncooperative members

    • Participation of active IESG and IAB members

    • Scope of charter
      • Fill existing slots

      • Stimulate reorganization or realignment

    Standards Process

    • Working pretty well

    • Some gray area end runs, e.g. 1597

    • Too creaky?

    • Too much freedom for splinter protocols?

    IESG Issues

    • Who defines areas?
      • Does IAB have a role?
      • IPng was defined and filled internally
    • What to do with vacancies?
      • App AD slot was filled by invitation
    • Basic issue is flexibility versus accountability
      • Possible additional issue is opportunity to serve

    IETF Charter

    • What does IETF do?
      • Standardization
      • Fresh Design


    • IRTF role and process is undefined
      • Not sure whether it matters
      • Not part of standards process
      • Unclear what comunity’s expectations are for RGs
      • What impact on resources?
    • IAB role is not completely clear
      • Appeals path for process problems
      • External representation,e.g. ISO Liaison
      • “Architecture” is vague; formal role undefined

    Suggested NomComm Changes

    • Chair
      • Remain non-voting except to resolve ties
      • Authority to handle whatever issues arise
      • Authority to ask IAB, IESG and IRTF to supply alternative representation.
    • Members
      • No alternates
      • 10 members
      • A quorum is 7
      • No member may be nominated to IESG or IAB
      • No sitting IESG or IAB member may serve except as liaison

    Other Unresolved Issues

    • What is the status of areas?
      • Formally defined and reviewed or IESG’s prerogative?
      • Does NomComm fill specific slots, does it define slots, or does it simply fill general positions?
    • How to deal with externally developed specifications

    • IETF Charter needs to be started


    • And many more…


        IAB Information Infrastructure Workshop

        John Romkey

        Internet Architecture Board

    Why a workshop?

    • Requested by the IESG

    • Work on common architectural issues in the Internet’s Information Infrastructure

    • Facilitate communication between communities


    • Date chosen: October 12-14
    • Location: probably Washington, DC
    • Initial invitations sent
    • Call for white papers issued

    Call for White Papers

      [See the minutes of the meeting.]


    • Understand the information infrastructure’s requirements of the lower layers

    • Understand the Internet’s requirements of the information infrastructure

    • Understand components of the information infrastructure

    • State succinctly the services that applications implementors may use

    Architectural Issues

      Laundry List:

      • naming
      • caching
      • type conversion
      • security issues (privacy, authorization, authentication, …)
      • replication
      • resource location
      • integration with directory services
      • accounting
      • encoding
      • longevity


    • Selected experts from inside and outside IETF
    • Selected white paper authors
    • IESG AD’s – Applications, User Services, Security
    • IAB members

    Next Steps

    • Publish workshop results in an Informational RFC
    • Publish results on-line
    • Report on workshop results at December IETF
    • Spread results beyond workshop

These minutes were prepared by Abel Weinrib, An online copy of these and other minutes are available in the directory