rfc1607.txt

RFC 的详细文档！


RFC 1607              A View from the 21st Century          1 April 1994


   "Collaboratories" which was intended to convey the idea
   that people and computers could carry out various kinds of
   collaborative work if they had the right kinds of networks
   to link their computer systems and the right kinds of
   applications to deal with distributed applications. Of
   course, we take that sort of thing for granted now, but it
   was new and often complicated 30 years ago.

   I am going to try to find out how they dealt with the
   problem of explosive growth.

   Louis and I will be leaving shortly for a three-day
   excursion to the new vari-grav habitat but I will let you
   know what I find out about the 1990s period in Internet
   history when we get back.

   Therese


   -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-


   To: "Troisema" <[e-mail]>
   CC: "David Kenter" <[e-mail]>
   CC: "Therese Troisema" <[e-mail]>
   From: "Jonathan Bradel" <[e-mail]>
   Date: September 13, 2023 10:34:05 LT
   Subject: Re: Internet History

   Therese,

   I sent a few Knowbot programs out looking for Internet
   background and found an interesting archive at the Postel
   Historical Institute in Pacific Palisades, California.
   These folks have an incredible collection of old documents,
   some of them actually still on paper, dating as far back as
   1962! This stuff gets addicting after a while.

   Postel apparently edited a series of reports called
   "Request for Comments" or "RFC" for short. These seem to be
   one of the principal means by which the technology of the
   Internet has been documented, and also, as nearly as I can
   tell, a lot of its culture. The Institute also has a
   phenomenal archive of electronic mail going back to about
   1970 (do you believe it? Email from over 50 years ago!). I
   don't have time to set up a really good automatic analysis
   of the contents, but I did leave a couple of Knowbots
   running to find things related to growth, scaling, and



Cerf                                                            [Page 8]

RFC 1607              A View from the 21st Century          1 April 1994


   increased capacity of the Internet.

   It turns out that the technical committee called the
   Internet Engineering Task Force was very pre-occupied in
   the 1991-1994 period with the whole problem of
   accommodating exponential growth in the size of the
   Internet. They had a bunch of different options for re-
   placing the then-existing IP layer with something that
   could support a larger address space. There were a lot of
   arguments about how soon they would run out of addresses
   and a lot of uncertainty about how much functionality to
   add on while solving the primary growth problem. Some folks
   thought the scaling problem was so critical that it should
   take priority while others thought there was still some
   time and that new functionality would help motivate the
   massive effort needed to replace the then-current version 4
   IP.

   As it happens, they were able to achieve multiple
   objectives, as we now know. They found a way to increase
   the space for identifying logical end-points in the system
   as well increasing the address space needed to identify
   physical end-points. That gave them a hook on which to base
   the mobile, dynamic addressing capability that we now rely
   on so heavily in the Internet. According to the notes I
   have seen, they were also experimenting with new kinds of
   applications that required different kinds of service than
   the usual "best efforts" they were able to obtain from the
   conventional router systems.

   I found an absolutely hilarious "packet video clip" in one
   of the archives. It's a black-and-white, 6 frame per second
   shot of some guy taking off his coat, shirt and tie at one
   of the engineering committee meetings. His T-shirt says "IP
   on everything" which must have been some kind of slogan for
   Internet expansion back then. Right at the end, some big
   bearded guy comes up and stuffs some paper money in the
   other guy's waistband. Apparently, there are quite a few
   other archives of the early packet video squirreled away at
   the PHI. I can't believe how primitive all this stuff
   looks. I have attached a sample for you to enjoy. They
   didn't have TDV back then, so you can't move the point of
   view around the room or anything. You just have to watch
   the figures move jerkily across the screen.

   You can dig into this stuff if you send a Knowbot program
   to [e-mail]. This Postel character must
   have never thrown anything away!!



Cerf                                                            [Page 9]

RFC 1607              A View from the 21st Century          1 April 1994


   Jon


   -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-


   To: "Jonathan Bradel" <[e-mail]>
   CC: "David Kenter" <[e-mail]>
   CC: "Troisema" <[e-mail]>
   From: "Therese Troisema" <[e-mail]>
   Date: September 15, 2023 07:55:45 UT
   Subject: Re: Internet History


   Jon,

   thanks for the pointer. I pulled up a lot of very useful
   material from PHI. You're right, they did manage to solve a
   lot of problems at once with the new IP. Once they got the
   bugs out of the prototype implementations, it spread very
   quickly from the transit service companies outward towards
   all the host computers in the system. I also discovered
   that they were doing research on primitive gigabit-per-
   second networks at that same general time. They had been
   relying on unbelievably slow transmission systems around
   100 megabits-per-second and below. Can you imagine how long
   it would take to send a typical 3DV image at those glacial
   speeds?

   According to the notes I found, a lot of the wide-area
   system was moved over to operate on top of something they
   called Asynchronous Transfer Mode Cell Switching or ATM for
   short. Towards the end of the decade, they managed to get
   end to end transfer rates on the order of a gigabyte per
   second which was fairly respectable, given the technology
   they had at the time. Of course, the telecommunications
   business had been turned totally upside down in the process
   of getting to that point.

   It used to be the case that broadcast and cable television,
   telephone and publishing were different businesses. In some
   countries, television and telephone were monopolies
   operated by the government or operated in the private
   sector with government regulation. That started changing
   drastically as the 1990s unfolded, especially in the United
   States where telephone companies bought cable companies,
   publishers owned various communication companies and it got
   to be very hard to figure out just what kind of company it



Cerf                                                           [Page 10]

RFC 1607              A View from the 21st Century          1 April 1994


   was that should or could be regulated. There grew up an
   amazing number of competing ways to deliver information in
   digital form. The same company might offer a variety of
   information and communication services.

   With regard to the Internet, it was possible to reach it
   through mobile digital radio, satellite, conventional wire
   line access (quaintly called "dial-up") using Integrated
   Services Digital Networking, specially-designed modems,
   special data services on television cable, and new fiber-
   based services that eventually made it even into
   residential settings. All the bulletin board systems got
   connected to the Internet and surprised everyone, including
   themselves, when the linkage created a new kind of
   publishing environment in which authors took direct re-
   sponsibility for making their work accessible.

   Interestingly, this didn't do away either with the need for
   traditional publishers, who filter and evaluate material
   prior to publication, nor for a continuing interest in
   paper and CD-ROM. As display technology got better and more
   portable, though, paper became much more of a specialty
   item. Most documents were published on-line or on high-
   density digital storage media.  The basic publishing
   process retained a heavy emphasis on editorial selection,
   but the mechanics shifted largely in the direction of the
   author - with help from experts in layout and
   accessibility. Of course, it helped to have a universal
   reference numbering plan which allowed authors to register
   documents in permanent archives. References could be made
   to these from any other on-line context and the documents
   retrieved readily, possiblyat some cost for copying rights.

   By the end of the decade, "multimedia" was no longer a
   buzz-word but a normal way of preparing and presenting
   information. One unexpected angle: multimedia had been
   thought to be confined to presentation in visual and
   audible forms for human consumption, but it turned out that
   including computers as senders and recipients of these
   messages allowed them to use the digital email medium as an
   enabling technology for deferred, inter-computer
   interaction.

   Just based on what I have been reading, one of the toughest
   technical problems was finding good standards to represent
   all these different modalities. Copyright questions, which
   had been thought to be what they called "show-stoppers,"
   turned out to be susceptible to largely-established case



Cerf                                                           [Page 11]

RFC 1607              A View from the 21st Century          1 April 1994


   law. Abusing access to digital information was impeded in
   large degree by wrapping publications in software shields,
   but in the end, abuses were still possible and abusers were
   prosecuted.

   On the policy side, there was a strong need to apply
   cryptography for authentication and for privacy. This was a
   big struggle for many governments, including ours here in
   France,  where there are very strong views and laws on this
   subject, but ultimately, the need for commonality on a
   global basis outweighed many of the considerations that
   inhibited the use of this valuable technology.

   Well, that takes us up to about 20 years ago, which still
   seems a far cry from our current state of technology. With
   over a billion computers in the system and most of the
   populations of information-intensive countries fully
   linked, some of the more technically-astute back at the
   turn of the millennium may have had some inkling of what
   was in store for the next two decades.

   Therese

   -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-


   To: "Therese Troisema" <[e-mail]>
   CC: "Jonathan Bradel" <[e-mail]>
   From: "David Kenter" <[e-mail]>
   Date: September 17, 2023 06:43:13 MT
   Subject: Re: Internet History

   Therese and Jon,

   This is really fascinating! I found some more material,
   thanks to the Internet Society, which summarizes the
   technical developments over the last 20 years. Apparently
   one of the key events was the development of all-optical
   transmission, switching and computing in a cost-effective
   way.  For a long time, this technology involved rather
   bulky equipment - some of the early 3DV clips from 2000-
   2005 showed rooms full of gear required to steer beams
   around. A very interesting combination of fiber optics and
   three-dimensional electro-optical integrated circuits
   collapsed a lot of this to sizes more like what we are
   accustomed to today. Using pico- and femto- molecular
   fabrication methods, it has been possible to build very
   compact, extremely high speed computing and communication



Cerf                                                           [Page 12]

RFC 1607              A View from the 21st Century          1 April 1994


   devices.

   I guess those guys at Xerox PARC who imagined that there
   might be hundreds of millions of computers in the world,
   hundreds or even thousands of them for each person, would
   be pleased to see how clear their vision was. The only
   really bad thing, as I see it, is that those guys who were
   trying to figure out how to deal with Internet expansion
   really blew it when they picked a measly 64 bit address
   space. I hear we are running really tight again. I wonder
   why they didn't have enough sense just to allocate at least
   1024 bits to make sure we'd have enough room for the
   obvious applications we can see we want, now?


   David


   -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-


Final Comments

   The letters end here, so we are left to speculate about many of the
   loose ends not tied up in this informal exchange. Obviously, our
   current struggles ultimately will be resolved and a very different,
   information-intensive world will evolve from the present. There are a
   great many policy, technical and economic questions that remain to be
   answered to guide our progress towards the environment described in
   part in these messages. It will be an interesting two or three
   decades ahead!




















Cerf                                                           [Page 13]

RFC 1607              A View from the 21st Century          1 April 1994


Security Considerations

   Security issues are not discussed in this memo.

Author's Address

   Vinton Cerf
   President, Internet Society
   12020 Sunrise Valley Drive, Suite 270
   Reston, VA 22091

   EMail: [phone]
   Fax: [phone]
   EMail: [e-mail]

   or

   Vinton Cerf
   Sr. VP Data Architecture
   MCI Data Services Division
   2100 Reston Parkway, Room 6001
   Reston, VA 22091

   Phone: [phone]
   Fax: [phone]
   EMail: [e-mail]

























Cerf                                                           [Page 14]