[ih] History from 1960s to 2025

Thu Dec 25 13:18:28 PST 2025

Craig,

That's very interesting. A few questions:

1. When exactly were I-Ds invented? (I know that it was no later than September 1989, see #3 below.)

2. Is it true to say that the de facto standard tool for producing early I-Ds was nroff?

(In 1994 when I wrote my first I-D, somebody -- very likely Scott Bradner -- sent me an nroff template, and I went on using it until XML2RFC first appeared.)

3. Who invented the formal expiry for I-Ds?

(It was first documented in RFC 1120 (Sept 1989) as far as I can tell, except that it was 3 months then, updated to "3-6 months" in RFC 1160, and codified as 6 months in RFC 1310.)

Incidentally, I think that RFC 1120 must have been the first RFC that documented the IETF standards process in any way.

Regards/Ngā mihi
    Brian Carpenter

On 26-Dec-25 09:19, Craig Partridge via Internet-history wrote:
> Hey Matt:
> 
> Your note brought up a flood of memories about creating the Internet-Draft
> series.  I don't think I've seen a history of the series and its creation,
> so I thought I'd dump my memories -- combined with some fact checking in
> IETF reports.
> 
> Very quickly in the IETF's development, it became clear that it was
> generating a large number of *interim* technical documents.  E.g. before
> each IETF meeting, a WG would typically produce a "latest draft" of
> whatever specification(s) it was working on, so they could be discussed.
> People wanted those drafts in a central spot, rather than just mailed to a
> WG, so they could figure out which WG meetings were the highest priority to
> attend during the IETF week.  Also, some documents were becoming big (100s
> of pages), an issue in a time of small disks which limited the size of
> emails.  So, it became clear a document series/repository/something else
> was needed.
> 
> As I recall, the initial ideas bounced around were to use the RFC series or
> revive IENs (Internet Engineering Notes so a logical series for the
> Internet Enginering Task Force).  Both were swiftly shot down.  Jon Postel
> and Joyce Reynolds did not want to place a flood of often partial-drafts of
> technical specs into the RFC series, nor deal with the tight timeframes
> (e.g. dozens of specs that all had to be published showing up a week before
> IETF meetings). For whatever reason, IENs were also declared off limits.
> 
> So Phill Gross, as chair of IETF, created a document series called IDEAS
> (announced at IETF 8 in NCAR in late 1987).  This produced pushback [my
> recollection here].  People wanted the IETF drafts to be ephemeral (fear
> that people would start claiming conformance to IDEA ### rather than RFCs,
> etc) and various other issues (which I only recall vaguely -- one issue, I
> believe, was the IAB was concerned this had the potential to end-run RFCs
> [see Note]).  As I recall, intellectual property issues were barely touched
> on. People realized things were being invented in WG meetings, but
> documenting them for posterity was not yet uppermost in folks' thoughts --
> thus the notion IETF documents could be ephemeral and would expire.
> 
> As late as IETF 11 (Ann Arbor, late 1988), there was still no document
> series in place -- IDEAS were sorta there (about a dozen ever existed), but
> not quite.  I note that Karen Bowers, a no nonsense, ex-military (?) person
> was brought in to manage many aspects of IETF including its documents
> around the time of IETF 11.  The fact that a year had passed and there was
> still no solution tells you the level of background discussions about how
> to create the needed document series.  Indeed, the cover note in IETF 11
> says, essentially, if you want to figure out where a WG is on its drafting
> of spec, contact Karen (!?!?!).  Remembering Karen's attitude on ad-hoc
> processes, I suspect she put some pressure on Phill and others to find a
> better answer ASAP.
> 
> Then at IETF 12 (January 1989) the Internet-Drafts series was announced.
> It has many of the elements of today's series; standard names, expiration
> after 6 months, draft plastered all over the document, in a form that can
> easily become an RFC.
> 
> Craig
> 
> Note re: RFCs.  It is worth remembering that just as IETF was spinning up
> (and the workload was quite big -- IETF 11's proceedings lists 10 active
> specs for things like Host Requirements, the first MIB, PPP, OSPF and an
> EGP successor) the RFC series was sputtering.  It produced about 25 RFCs in
> 1986 and a similar number in 1987.  It was clear the IETF was going to more
> than double that annual total -- in other words, IETF product would soon
> dominate the RFC series. The IAB (and Jon P) wanted to retain control of
> RFCs and protocols deemed part of the Internet architecture.  This created
> a potential dilemma - if the IETF created its own document series, so RFCs
> only saw final versions of specifications, that met Jon's need to not
> publish ephemeral stuff, but raised the possibility that the IETF could
> weaponize its document series to undermine RFCs if specs did not mature to
> RFC status after IETF felt they were ready.
> 
> On Thu, Dec 25, 2025 at 9:58 AM Matt Mathis via Internet-history <
> internet-history at elists.isoc.org> wrote:
> 
>> One key development (that predates me, so I can't provide details) was the
>> codification (and evolution) of the Internet Draft and RFC processes.    I
>> believe that finding the right balance between ease of contribution,
>> permanence and implied or explicit (non)authority, embodied by the use of
>> the name "Request For Comments" was as important as any individual
>> technical detail.   The publication process substantially inspired the
>> culture of the IETF (or perhaps vice-versa), which is what enabled
>> collaborative engineering between nominally competing organizations.
>>
>> As far as I know RFCs were the first ever self published archival series of
>> documents.
>>
>> Thanks,
>> --MM--
>> Evil is defined by mortals who think they know "The Truth" and use force to
>> apply it to others.
>> -------------------------------------------
>> Matt Mathis  (Email is best)
>> Home & mobile: 412-654-7529 please leave a message if you must call.
>>
>>
>>
>> On Sat, Dec 20, 2025 at 6:09 PM Karl Auerbach via Internet-history <
>> internet-history at elists.isoc.org> wrote:
>>
>>>
>>> On 12/18/25 12:21 PM, John Day via Internet-history wrote:
>>>> And some of us thought, it was the continuation of building a
>>> resource-sharing network.  ;-)
>>>
>>> In the mid 1980's I spent a year or more at the Livermore Labs working
>>> on the MFE (magnetic confinement fusion energy) project. (Playing tennis
>>> with a multi-million degree ball of plasma as the ball was kinda fun.)
>>>
>>> I wasn't involved in the networking part but I certainly overheard a lot
>>> of expressed desire to share not only our simulations and measurements
>>> (we had a couple of seriously-gigantic fusion vessels across the road
>>> from my office) as well as our boatload of Cray machines and data
>>> libraries.
>>>
>>> The folks at the labs were pretty good a jury rigging things and it is
>>> my understanding that they created some duct-tape-and-bailing-wire
>>> systems to do that kind of sharing.
>>>
>>> Also, in the 1970's when I was at SDC I heard many tales about the Q7
>>> and Q32 computers, and the desire to time share the latter among
>>> research institutions.  But I have no real memory of what was said in
>>> those tales.
>>>
>>>       --karl--
>>>
>>>
>>> --
>>> 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
>>
> 
> 