This is only a rough draft - Megan 04/10/92 23rd IETF meeting -- San Diego, CA Working Group on Internet School Networking March 17, 1992 1:30 -- 3:30 p.m. Co-Chairs: Art St.George (present), stgeorge@bootes.unm.edu; Connie Stout, connie.stout@tenet.edu; John Clement (present), clement@educom.edu General mailing list for ISN-WG: isn-wg@unmvm.unm.edu A list of those attending the session is appended to this document. Minutes Specific discussion around the major agenda topics is summarized below in sections 1) and 2). A number of general issues were addressed during the meeting, and are summarized briefly first. Joyce Reynolds mentioned that there is considerable interest in international circles, especially in Europe, in what this working group is doing, and that products from the WG will be scrutinized attentively overseas. Among the more general issues (exceeding technical boundaries) raised at least briefly during the session were: the need to provide persuasive arguments for educators to use with school administrators, to support both initial access and expansion of networking capabilities; the continuing need for postsecondary institutions to remain involved in helping schools with connectivity, whether as sources of guest accounts, for technical expertise and support where other mechanisms are not yet in place, or as sources of collaboration on connectivity or content problems. The issue of having the ISN-WG contribute to providing technical consulting capabilities to K-12 groups and institutions was discussed briefly. Although no definite resolution was offered, after the meeting it was suggested that a roster of persons willing to respond to technical questions might be prepared and offered as part of the FAQ resource (cf. item 2) below), with the possibility that these individuals might provide further technical consultation. One condition of listing in the roster would be that good questions to roster members would be added by them to the FAQ list. Denis Newman offered that there are many more (both local- and wid-area) networks in place in schools for administrative uses than there are instructional- use networks. He stipulated that, if administrative networks can be linked into the Internet -- as FIRN is in Florida -- overall connectivity can be substantially advanced. Denis suggested that the working group examine the issues involved in extending administrative LANs and WANs for instructional use; in particular, the group might consider addressing security questions, and look at the reality of risks entailed in carrying administrative and instructional traffic over the same networks, especially ways to minimize the risk of unauthorized access. 1) Review of connectivity alternatives and growth paths. Mailing list for this discussion: connect@unmvm.unm.edu A draft document prepared by Pat Burns and Ed Zachmann of Colorado State University (available on connect archives in Postscript format, or contact Art St.George for fax) was distributed to attendees, and served as a basis for extensive discussion during the meeting. John Clement presented a networking growth path from the viewpoint of educator practices. Ensuing discussion pointed out that the Burns and Zachmann paper lacked an initial connection model, what might be called Level 0 (the present paper begins with Level 1). Burns and Zachmann's Level 1 connectivity models (cf. especially Figures 3a and 3b) led to a consensus that both figures and their explanatory text needed relatively minor but significant revisions. Extensive discussion of the need for continued connection capability for the existing base of computers in schools (often machines of reduced capacity) led to an expressed consensus that such capability should be maintained in the near future (under Level 0 connection options), but should be eliminated over time as more advanced connection levels predominate. The minimum hardware capability for full Internet access to be supported would be Apple Macintosh machines or MS- DOS machines running Windows. Joe Blackmon summarized his experiences in putting together full (56kbps) Internet connections for finalist schools in the SuperQuest competition at lowest possible costs, and offered to share a document he is preparing for SuperQuest on the subject. Discussion of Levels 2-4 (Figures 4-6 of Burns and Zachmann paper) connection models was relatively brief. There seemed to be general agreement that, on first reaction (most attendants were seeing the paper for the first time at the session), the models were acceptable with relatively minor corrections, although considerable extension and clarification were requested (see below). One point made was that the models as presented in the paper were overspecified with regard to their technical components, for the level of generality needed for a paper on connectivity alternatives. At the very least, mention should be made of other technical alternatives. [This was in no way presented as a criticism of the Burns and Zachmann paper, since they offered a starting point and it was appropriate to list a complete technical implementation so as to estimate costs. But perhaps it could be presented as one example.] Among the points raised that implied expanding the descriptions of the models: o guidance on the boundaries of application of each model: when each becomes inappropriate in given school contexts. For example, model 1 might apply to 3-8 machines in a school, and would not work for more than eight machines. o specification of what kinds of Internet capabilities are available for each model: i.e., model 1 offers telnet (remote login) via an interface menu item, but no direct telnet capability, and would allow access to WAIS via Simple WAIS but not the installation of WAIS client software, etc. o more complete specification of what is required at the Internet node/connection point for each model; o cost estimates for moving from one model to the next, and stipulation of what equipment would be no longer needed and could be used elsewhere; o specification of personnel, time and training requirements for installation, support (including time for administration and security protection), and maintenance. Jeff Hayward (jhayward@utexas.edu--ART I'M NOT SURE THIS ADDRESS IS RIGHT, DOING IT FROM MEMORY AND WHOIS IS NO HELP..) agreed to lead an effort to respecify the models on the basis of the Burns and Zachmann paper and its discussion, and to amplify their written description, and to prepare a draft before the end of the IETF meeting if possible. It was agreed that a revised draft document would be made available on , and that it would be revised once more from feedback on the list. A twice- iterated draft might be available in advance of the Twenty- Fourth IETF, in Boston in July. 2) Development of an FAQ ("frequently-asked-questions) archive on school connectivity issues. A new mailing list will be formed for this issue Discussion initially focused on identifying sources of information and materials from which a set of FAQs and candidate answers might be developed. KIDSNET and EDTECH lists were mentioned, and there are many others. A number of attendees offered to help round up relevant materials: Gene Hastings Peter Deutsch Rob Reilly Art St. George and John Clement offered to assemble the materials located into a preliminary archive. Volunteers were solicited to edit the archive and review/revise answers to the questions. The following were dragooned into volunteering: Tony Rutkowski Joe Blackmon Tracy LaQuey Michael Marcinkevicz John Clement will also ask Al Rogers of FrEdMail to volunteer or suggest one of the FrEdMail sysops to volunteer. By the next IETF meeting, it was agreed that a preliminary FAQ archive would have received an editing pass and would be available for examination and revision. The following preliminary set of categories is offered for consideration for the FAQ list: o why should K-12 educators and schools connect to the Internet? o modes of access o costs o support issues, including maintenance, access rules, security o sources of information o troubleshooting o glossary of essential terms o technical specifications o technical consultants roster: people willing to respond to questions, or possibly provide further consultation