box@mamnix.as.sub.org (Fido Daemon) (05/26/90)
A great unified net is almost impossible! Every net uses its own software, every net has another mail time. How can it be managed to make 10,000 or more BBSes to switch to a new, "Great Unified" network from today to tomorrow? This would be so much administra- tion work... Fabian <DX-110-15000>
schoff@uu.psi.com (Martin Schoffstall) (05/27/90)
>"The syntax required to send a message from one machine to >another is often complicated, and there is no central register of electronnic >addresses for users... The current state of computer networks is somewhat akin >to telephone systems around the turn of the century - there are numerous >systems, some mutually incompatible, and some interconnected..." > There is substantial momentum today to conform to domain notation of the form someuser@somehost.someplace.somecountry where "somecountry" is an official ISO country code. This helps alot. > >The Guide then gives many helpful hints how to overcome communications >problems, and at the end it sounds an encouraging message: "Hopefully, in time, >the current complexities of electronic mail will disappear and something like >the relative simplicity of the national and international telephone systems >will replace it." > In conjunction with other organizations in the UK, Sweden, etc.. PSI has tried to take the X.500 Directory and between some administrative discipline and a pilot schema created a WhitePages service. In the US that has brought 50 organizations and 300K records into the disribute system, looking for "Martin Schoffstall" and ignoring the infrastructure information of "somehost.someplace.somecountry" seems possible. > >Really? Are there concrete plans to abandon all computer networks for the Great >Unified Net? I've little knowledge about the people/organizations behind all >these nets - would there be a 'central authority' that could enforce such a >unification? > I'd posit that there is a grass roots effort to make the Internet be the "Unified Net" without a central authority, but based on market pressures. Marty
schoff@uu.psi.com (Martin Schoffstall) (05/27/90)
Donald, While I agree with much of what your saying on a philosphical level 1) tcp/ip has delivered 2) The Domain notation is prevalent and growing 3) encourage your organization to participate in the Internet etc... [ I've already posted I need to argue with you on some fine points: >Internet is that networking specialists all over the world are under >pressure to deliver effective interoperable systems *NOW*, and the >TCP/IP suite of protocols used by the Internet is the only proven >technology that can do the job. OSI/X.400/X.500 is simply not ready >for worldwide production use yet. In country after country during the The International Standards Organization (ISO) clearly has a string of failures in FTAM, VT, TP0, etc... However, there is an element of the tcp/ip community that thinks that X.400 and X.500 are worthy enough to exist as applications in the tcp/ip world. As I've already mentioned PSI and others have succesfully been using it over tcp/ip on a large scale. Certain places (like Columbia University) use X.500 as the basis for distributing their phonebook data on a timely basis (weekly I believe), to faculty/student/staff who participate in the Columbia wide network. It is possible that we will hide much of the ugliness of the X.400 notation but use the protocol mechanisms to deliver multimedia mail, again, layered on tcp/ip. Marty
dwells@fits.acc.Virginia.EDU (Don Wells) (05/27/90)
Marty, In article <1990May26.190406.7788@uu.psi.com> schoff@uu.psi.com (Martin Schoffstall) writes: >I need to argue with you on some fine points: As you will see below, I am in complete agreement with your fine points. >The International Standards Organization (ISO) clearly has a string >of failures in FTAM, VT, TP0, etc... However, there is an element of >the tcp/ip community that thinks that X.400 and X.500 are worthy >enough to exist as applications in the tcp/ip world. Count me as a part of that element. I loosely referred to "OSI/X.400/X.500" because the previous followup had argued for the X.400 address notation; my usage was misleading, because what I really meant was the effort to deliver an entire interoperable OSI protocol stack, which I agree has been a limited success at best. > Certain places (like Columbia University) use X.500 >as the basis for distributing their phonebook data on a timely basis >(weekly I believe), to faculty/student/staff who participate in the >Columbia wide network. I am aware of the NYSERnet/PSI pilot project, and am watching it closely. I am on the "QUIPU" mailing list [*], so you know I am seriously interested in X.500. I, wearing my hat as networking coordinator for NRAO, judge that your project is still somewhat a beta trial, and that NRAO's role is to be on the cutting edge of astronomy research, not to be on the cutting edge of CS research, and therefore I wait and watch. I am impressed by what I hear of your pilot, and my present guess is that I might recommend that we participate in a year or so. >It is possible that we will hide much of the ugliness of the X.400 >notation but use the protocol mechanisms to deliver multimedia mail, >again, layered on tcp/ip. Hear! Hear! I definitely want that elegant multimedia functionality as soon as I can get it for our Suns, Convexes, etc., at about the same price that I pay for the SMTP mailers and the DNS servers and resolvers that we now use (i.e., at no extra charge). Obviously X.400_over_TCP/IP and X.500_lookup will coexist with good old SMTP mail and the DNS for many years, so that we can have a graceful, relatively painless transition to the brave new world. I agree that the address notation currently proposed for X.400 mailers is really ugly compared to the DNS notation. IMO, finishing the worldwide Internet ASAP and then carefully transitioning it to the new application protocol suites, while maintaining interoperability, is now the only rational strategy open for worldwide networking. This is why I urged my fellow astronomers to connect to the Internet ASAP so that they can "have it all" right now: real interoperability plus the distributed DNS database plus a realistic migration strategy to the OSI application layer components. I judge from your remarks that you and I are in very nearly perfect agreement on all these issues. Best regards, Don Donald C. Wells, Associate Scientist | NSFnet: dwells@nrao.edu [192.33.115.2] National Radio Astronomy Observatory | SPAN: NRAO::DWELLS [6654::] Edgemont Road | BITnet: DWELLS@NRAO Charlottesville, VA 22903-2475 USA | UUCP: ...!uunet!nrao.edu!dwells Lat: 38:02.2N Long: 78:31.1W | Tel:+1-804-296-0277 Fax:+1-804-296-0278 [*] Note to readers interested in the X.500 technology: quipu@cs.ucl.ac.uk is a worldwide exploder for discussions of issues surrounding another X.500 prototype effort called QUIPU. Send requests to quipu-request@cs.ucl.ac.uk if you want to listen. I recommend that only people *seriously* interested in details of X.500 technology listen to this list. If you want to participate in the NYSERnet/PSI pilot send a request for info to Marty (schoff@uu.psi.com). Again, only do this if you are ready to do a lot of work (anybody who has tried to construct and maintain an organization-wide phone directory can tell you it is a major project). You would also want to listen to several relevant newsgroups (e.g., comp.protocols.iso.x400 and comp.protocols.tcp-ip.domains), and probably like me be baffled by much of the arcane semi-gibberish ISO/OSI notations.
schoff@uu.psi.com (Martin Schoffstall) (05/28/90)
>I am aware of the NYSERnet/PSI pilot project, and am watching it >closely. I am on the "QUIPU" mailing list [*], so you know I am >seriously interested in X.500. I, wearing my hat as networking >coordinator for NRAO, judge that your project is still somewhat a beta >trial, and that NRAO's role is to be on the cutting edge of astronomy >research, not to be on the cutting edge of CS research, and therefore >I wait and watch. I am impressed by what I hear of your pilot, and my >present guess is that I might recommend that we participate in a year >or so. > While you are correct that this is CS research issue, the customers of NYSERNet/PSINet have had the possability to opt for simply handing us the data and letting us take care of the hard parts. We're hoping sometime this year to go from a pilot to service status (at least for our customers), this is going to be a major point of discussion at the June meeting of PSITech (our technical users group). Marty
roberts@nimrod.wpd.sgi.com (roberts) (05/31/90)
In article <1990May26.182226.7277@uu.psi.com>, schoff@uu.psi.com (Martin Schoffstall) writes: > >Unified Net? I've little knowledge about the people/organizations behind all > >these nets - would there be a 'central authority' that could enforce such a > >unification? > > > I'd posit that there is a grass roots effort to make the Internet be the > "Unified Net" without a central authority, but based on market pressures. ^^^^^^^^^^^^^^^^^^^^^^^^^^^ ^^^^^^^^^^^^^^^^^^^^^^^^^ Down with centralized planning and authority! The free market rules! Resist the efforts of totalitarian bureaucracies to regain control of the information economy. - Robert Stephens