[ih] Governments' Bureaucracy and The Internet
Jack Haverty
jack at 3kitty.org
Sat Nov 7 13:57:22 PST 2026
In the 1980s at BBN, there were other IMP networks that we installed and
sometimes also operated for commercial clients. One of them we operated
had nodes in all sorts of weird places. One weird place was some small
country in South America. We encountered a similar problem with
ordering lines. The nation's "phone company" had a 9-month delay to get
a new line. Maybe. If they got around to it. Don't plan on it.
The client had a simple solution to the problem. They simply bought
the local phone company. Problem solved.
/Jack
On 11/7/26 13:46, Steve Crocker wrote:
> I can't speak about the role of the bureaucracy in creating the
> Internet. I saw a tiny bit of the bureaucracy during the creation of
> the Arpanet.
>
> Two particularly relevant pieces of the bureaucracy related to the
> procurement of the 50 kilobit per second leased lines. On the
> positive side, the government had access to a special price, I believe
> that made it possible use the lines instead of much slower lines and
> stay within the available budget.
>
> On the other hand, the rules for ordering the lines included a nine
> month lead time. That made it very hard to adjust to the occasional
> issues that arose requiring replanning the topology.
>
> Steve
>
>
> Sent by a Verified
>
> sender
>
> On Sat, Nov 7, 2026 at 4:35 PM Jack Haverty <jack at 3kitty.org> wrote:
>
> This list reminds me a lot of the Internet Meetings more than 40
> years ago -- whatever the current topic was, it was easy to go off
> on tangents that could only be corrected by activation of the
> Rathole Protocol (search the archives of this list).
>
> I changed the subject of this message to reflect the current
> distraction.
>
> In the BBN world, someone started putting a serial number in the
> Title of the report, e.g., "Quarterly Technical Report #23". That
> made it easier to track the QTRs, since the Title was also on the
> cover page. I just continued the practice and don't know who
> started it. Probably some experienced old guy.
>
> The lesson I learned was that it was far easier to work around a
> bureaucratic rule than to try to change it.
>
> At BBN we built hardware for the government. The contract
> boilerplate was often huge, containing all sorts of rules as
> dictated by laws. I recall one required the contractor (us) to
> procure all ball bearings used in the design from a particular
> company in some US state. Fortunately there is little use for
> ball bearings in computer equipment. But I've always wondered
> about the History of that rule and law that required it. Some
> legislator somewhere got something for his or her insistence on
> including that requirement in passing the law that led to the
> procurement regulation in the boilerplate.
>
> But the bureaucratic history is also an important, and likely
> influential, part of Internet History. Did bureaucracy help
> establish The Internet, or hinder it? There were other governments
> involved, not just the US. What kinds of such issues were there in
> Europe or elsewhere as the Internet grew?
>
> /Jack Haverty
>
> On 11/7/26 11:45, Steve Crocker wrote:
>> :) I had my own encounters with government bureaucracy. One day
>> I went to the file and pulled a few years' worth of QTRs. Each
>> one had a cover sheet -- not the DD-54, but a cover sheet
>> resembling a title page. The cover sheet contained all sorts of
>> information. Contract number, contractor name, contract dates,
>> total contract amount, title of the work, etc. However, none of
>> this information indicated the report date or sequence number or
>> anything that changed from one report to the next. Nor did the
>> content of each report help very much ;)
>>
>> Upon investigation, I discovered the contract contained explicit
>> language dictating the cover page content of each QTR, and the
>> QTRs were in complete conformance.
>>
>> I called the ARPA Program Management Office, our interface to the
>> contracting offices, and suggested they fix this. Nothing
>> happened. Some time later, I called them again. Nothing
>> happened. Yet later, completely exasperated, I wrote a memo
>> explaining the issue once again and mentioned that I had already
>> reported it twice before. This triggered a desperate search for
>> my previous two complaints. Since my previous two complaints
>> were verbal, there was no record.
>>
>> And in an entirely different setting, I once had a professor call
>> me asking for relief from a contract constraint. He had written
>> a lengthy report on their AI research and discovered an explicit
>> formula in the contract limiting the total number of pages to
>> 25,000 (i.e., number of pages x number of copies ≤ 25,000). He
>> wanted to print more copies. I thought it was a good idea since
>> reporting on AI research was skimpy in those days and I thought
>> it would help the cause. His contract was part of my budget, but
>> I was only the COTR. Relief from a contract requirement could
>> come only from the Contracting Officer. I called the Contracting
>> Officer. He was a really old guy, maybe 50 years old. I was in
>> my twenties. I explained the issue. He said no. I explained it
>> again and said it would be helpful. He said no. I got a little
>> agitated and inappropriately suggested we could move the next
>> contract renewal to another Contracting Office. He said we could
>> do whatever we wanted, but that constraint was a statute. A
>> statute! Not an OMB circular, not a FAR regulation, not an
>> Agency directive? There's a whole parfait of regulations and
>> rules. Statutes are at the base. That means Congress passed a law.
>>
>> He waited a couple of beats, either out of patience with my youth
>> or enjoying the moment. "Reports bigger than that have to be
>> sent to the Government Printing Office." The light dawns. I can
>> fill in the rest. And farmed out to various bidders. I thanked
>> him for his time, ended the call, and reorganized. I wasn't
>> about to give up. I could have asked someone else in the office
>> or figured it out myself. Instead, I decided to ask someone else
>> who might have faced this problem. I estimated the university
>> was roughly at the B+ level of sophistication in dealing with
>> government bureaucracy. I called a friend at the Rand
>> Corporation. Rand had fifty years of experience dealing with the
>> U.S. government. Almost certainly at the A+ level.
>>
>> My friend listened briefly and responded, "You mean it's a report
>> with appendices and is in multiple volumes and perhaps has been
>> through a couple of versions?" "Why, yes," I said, "that's
>> exactly what he has."
>>
>> I called the professor back, explained about appendices and
>> volumes, then sat back in my chair feeling very good about my
>> service to the community and to the people and government of the
>> United States.
>>
>> Many years later when I shared this episode with a one of the
>> program managers who had the same job, he cut me off quickly and
>> said simply cut it into multiple pieces. But he was quite a bit
>> older than I had been. Experience counts.
>>
>> Steve
>>
>>
>>
>>
>> On Sat, Nov 7, 2026 at 2:10 PM Jack Haverty via Internet-history
>> <internet-history at elists.isoc.org> wrote:
>>
>> Steve - your mention of late quarterly reports reminded me of
>> my first
>> encounter with government bureaucracy.
>>
>> Sometime in the late 1970s, I was PI on some ARPA contract
>> with BBN.
>> Our Finance guy accosted me one day to say that the
>> government was
>> refusing to pay a bill because we had not yet submitted our
>> Quarterly
>> Technical Report, aka QTR. I clearly remembered submitting
>> that report,
>> on time, after a struggle internally to get everybody
>> involved to write
>> their piece. After some investigation, the ARPA PM (may have
>> been Vint)
>> confirmed that ARPA had received the QTR. But the government
>> still
>> refused to pay the bill.
>>
>> With more investigation, we discovered that all Deliverables
>> had to be
>> accompanied by a government form (a "DD-54" IIRC). The form
>> for our
>> QTR had either been forgotten by the BBN bureaucracy or lost
>> in its
>> travels through the government bureaucracy. No DD-54, no
>> Deliverable
>> received. The actual QTR didn't really matter. Without the
>> DD-54, the
>> government withheld payment, but no one seemed able to
>> explain why.
>>
>> That was my first lesson in how to deal with government
>> bureaucracy -
>> just do what the system requires. I didn't know Jon had
>> worked on the
>> Keydata system (if it was actually in use yet at ARPA), or I
>> might have
>> called him up and reported a Bug. I suspect he would have
>> told me it's
>> a Feature.
>>
>> /Jack Haverty
>>
>> On 11/7/26 06:15, Steve Crocker via Internet-history wrote:
>> > Lyman,
>> >
>> > Thanks! I was about to send a follow up note re Keydata.
>> >
>> > The PDP-15 was for a data management system inside DARPA.
>> It held copies
>> > of the contracts and management reports. The system was a
>> one-off system
>> > developed by Keydata under contract through our office
>> (IPTO). It was an
>> > interactive, time-shared system, and the agency director,
>> Steve Lukasik,
>> > had direct access to it.
>> >
>> > Prior to the creation of that system, all of the tracking
>> of contract
>> > status, expenditures, etc were done through a DoD system at
>> Cameron
>> > Station, somewhere near the Pentagon. Paper reports
>> arrived monthly. In
>> > addition to changing the timeliness of reports, it also
>> disrupted the flow
>> > of control. One day I got a call from the director asking
>> why contractor X
>> > hadn't submitted any quarterly reports in the last few
>> years. Prior to the
>> > creation of the Keydata system, if the director had been
>> interested in past
>> > quarterly reports, he would have asked someone. In the
>> process of asking,
>> > we would have become aware of the situation and we would
>> likely have
>> > reached out to the contractor with an urgent request/demand
>> for updates.
>> >
>> > We had a TIP in the agency, so connecting the PDP-15 as a
>> host was a
>> > natural step. And Jon was just the right person to help
>> Keydata add
>> > support for NCP and Telnet to the Keydata operating system.
>> >
>> > Steve
>> >
>> >
>> > Sent by a Verified
>> >
>> > sender
>> >
>> > On Sat, Nov 7, 2026 at 8:46 AM Lyman Chapin via
>> Internet-history <
>> > internet-history at elists.isoc.org> wrote:
>> >
>> >> A quick search shows that the CV is also online at
>> >>
>> https://www.postel.org/wp-content/uploads/sites/32/2023/07/postel-cv-1997.txt
>> >>
>> >> - Lyman
>> >>
>> >>> On Nov 7, 2026, at 8:08 AM, vinton cerf via
>> Internet-history <
>> >> internet-history at elists.isoc.org> wrote:
>> >>> Can any of you help to recreate Jon Postel's work history?
>> >>>
>> >>> UCLA - 1967(?) - 1972 (?)
>> >>> SRI XXXX-XXXX
>> >>> <unknown>
>> >>> USC-ISI XXXX - 1998
>> >>>
>> >>> thanks
>> >>>
>> >>> vint
>> >>> --
>> >>> Internet-history mailing list
>> >>> Internet-history at elists.isoc.org
>> >>> https://elists.isoc.org/mailman/listinfo/internet-history
>> >>> -
>> >>> Unsubscribe:
>> >>
>> https://app.smartsheet.com/b/form/9b6ef0621638436ab0a9b23cb0668b0b?The%20list%20to%20be%20unsubscribed%20from=Internet-history
>> >>
>> >> --
>> >> Internet-history mailing list
>> >> Internet-history at elists.isoc.org
>> >> https://elists.isoc.org/mailman/listinfo/internet-history
>> >> -
>> >> Unsubscribe:
>> >>
>> https://app.smartsheet.com/b/form/9b6ef0621638436ab0a9b23cb0668b0b?The%20list%20to%20be%20unsubscribed%20from=Internet-history
>> >>
>>
>> --
>> Internet-history mailing list
>> Internet-history at elists.isoc.org
>> https://elists.isoc.org/mailman/listinfo/internet-history
>> -
>> Unsubscribe:
>> https://app.smartsheet.com/b/form/9b6ef0621638436ab0a9b23cb0668b0b?The%20list%20to%20be%20unsubscribed%20from=Internet-history
>>
>>
>>
>> --
>> Sent by a Verified
>>
>> sender
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: OpenPGP_signature.asc
Type: application/pgp-signature
Size: 665 bytes
Desc: OpenPGP digital signature
URL: <http://elists.isoc.org/pipermail/internet-history/attachments/20261107/de4a93af/attachment.asc>
More information about the Internet-history
mailing list