[ih] "The Internet runs on Proposed Standards"

[Brian E Carpenter]

Jack, this is perhaps just a parable, but I think it says a lot:

Soon after HTML and HTTP made their debuts, I attended a seminar
at CERN by Frank Kappe, a student of Hermann Maurer at TU Graz,
Austria, about the Hyper-G project. Their strongly expressed opinion
was that the Web was useless as a hypertext project because (unlike
Hyper-G) it did not have rigorous two-way hyperlinks. They maintained
that an unmanaged hypertext system in which A could point to B but
B didn't know about it and have a reverse pointer to A was useless.
How could it possibly be managed and kept consistent?

Needless to say, a certain member of the audience didn't agree. Tim
Berners-Lee argued that the two-way hyperlink approach didn't scale,
required some sort of coordinated database to work at all, and would
never succeed as a result. Whereas the Web would scale indefinitely,
needed no centralized anything, didn't need managing, and had a chance
of world domination.

This seminar was in September 1994. At that time there was only one
Informational RFC about the Web (RFC1630), which didn't even describe
HTML and HTTP. But the Web was already running in several countries
(this was probably around the time that the Web overtook Gopher).

So yes, not rigorous, not remotely standardised, still considered
an upstart by many people, certainly no IETF rough consensus, but
above all, running code.

Hyper-G??? Yes, a real project, later renamed Hyperwave, which you
can read about at https://mprove.de/visionreality/text/2.1.15_hyperg.html
World domination? Not so much.

The point of the parable is that rigorous, well defined, good computer
science and solid engineering simply haven't worked as well in the Internet
as moving fast and breaking stuff.

Regards
    Brian Carpenter

On 04-Dec-22 13:06, Jack Haverty via Internet-history wrote:
> Hi Andy,
> 
> Yes, that's a good example.  But what kind of example....?   If I were
> evaluating a proposal, I'd have some questions.
> 
> - Does anyone, other than the manufacturer's own marketing department,
> agree that the "standard" technology is implemented and operational in
> the product?
> - Has another manufacturer implemented their own independent
> implementation?  Has it been tested to interact properly with the first
> manufacturer's implementation?
> - Has any independent entity tested the implementations and verified it
> is operating as defined, and "certified" the product to be true to the
> Standards specification.
> - Has any early adopter integrated the new product into an existing
> operational network system?  How did they accomplish the changeover?
> 
> Depending on the answers, the example might be a successful deployment
> of an Internet Standard.   Or it might be an example of a proprietary
> solution that is only known to work in a network composed only of
> equipment from that manufacturer -- i.e., a "walled garden" technology
> which happens to use mechanisms defined in RFCs.
> 
> My point about "process" is the apparent lack of mechanisms analogous to
> things like UL for the electricity infrastructure, and rules and
> regulations such as building codes, and enforsement mechanisms such as
> inspectors - i.e., things typically associated with mature
> "infrastructures".
> 
> The "deployment process" of the Internet in the 1980s was crude and
> rudimentary, but it existed and was somewhat effective.  Somehow along
> the way it seems to have disappeared even as the technology evolved.
> Perhaps some Historian will explore that someday.
> 
> Jack
> 
> On 12/3/22 14:23, Andrew G. Malis wrote:
>> Jack,
>>
>> I've been on both the sending and receiving side of Internet Backbone
>> Equipment RFPs. All you need to do is take a look at the cited RFCs
>> and drafts (not even yet RFCs) and you'll see what I mean. I realize
>> that's not public information and thus hard to defend.
>>
>> However, you can easily take a look at the online documentation for
>> any big-iron router from Cisco or Juniper, or other router vendor.
>>
>> I just picked, at random, Cisco's "MPLS Configuration Guide for Cisco
>> 8000 Series Routers, IOS XR Release 7.5.x", which you can find at
>> https://www.cisco.com/c/en/us/td/docs/iosxr/cisco8000/mpls/75x/b-mpls-cg-cisco8000-75x/implementing-mpls-ldp-75x.html
>> . MPLS is a major feature in today's backbone routers, and goes back
>> to the early 2000s for first backbone deployments. If you scroll down
>> to the referenced RFCs at the end of the document, they are all
>> Proposed Drafts.
>>
>> There are many many other examples out there.
>>
>> Cheers,
>> Andy
>>
>>
>> On Sat, Dec 3, 2022 at 4:34 PM Tony Li <tony.li at tony.li> wrote:
>>
>>
>>
>>>      On Dec 3, 2022, at 12:37 PM, Jack Haverty via Internet-history
>>>      <internet-history at elists.isoc.org> wrote:
>>>
>>>      I still have doubts about "The Internet runs on Proposed
>>>      Standards".   Does anybody know -- Is it true?  How do you know?
>>>      Personally I haven't found any way, at least as a User, to tell
>>>      what technology is inside all the equipment, software, services,
>>>      protocols, algorithms, et al that are operating between my
>>>      keyboard/screen and yours.  It could be all Standards of some
>>>      ilk, or it could all be Proprietary.   It might conform to the
>>>      spec, or have some zero-day flaw.  How do you tell?
>>
>>
>>      Data point: BGP is still a draft standard.
>>
>>      Some might argue that the Internet cannot run without BGP.
>>
>>      I’m not quite convinced.  I suggest we turn it off and find out.
>>      It should be an interesting experiment.
>>
>>      Regards,
>>      Tony
>>