Analysing Internet Standards Development Organisation Data
draft-perkins-analysing-sdo-data-00
This document is an Internet-Draft (I-D).
Anyone may submit an I-D to the IETF.
This I-D is not endorsed by the IETF and has no formal standing in the
IETF standards process.
The information below is for an old version of the document.
| Document | Type |
This is an older version of an Internet-Draft whose latest revision state is "Active".
|
|
|---|---|---|---|
| Authors | Colin Perkins , Ignacio Castro , Ryo Yanagida , Stephen McQuistin | ||
| Last updated | 2026-05-06 | ||
| RFC stream | (None) | ||
| Formats | |||
| Stream | Stream state | (No stream defined) | |
| Consensus boilerplate | Unknown | ||
| RFC Editor Note | (None) | ||
| IESG | IESG state | I-D Exists | |
| Telechat date | (None) | ||
| Responsible AD | (None) | ||
| Send notices to | (None) |
draft-perkins-analysing-sdo-data-00
RASPRG C. Perkins
Internet-Draft University of Glasgow
Intended status: Informational I. Castro
Expires: 7 November 2026 Queen Mary University of London
R. Yanagida
S. McQuistin
University of St Andrews
6 May 2026
Analysing Internet Standards Development Organisation Data
draft-perkins-analysing-sdo-data-00
Abstract
This document outlines some issues to consider when studying data
relating to the Internet standards development ecosystem. It
identifies observable components of standards development processes,
proposes a taxonomy of possible measurements, and highlights
methodological, interpretive, and ethical considerations. It is
intended to support a range of uses, including monitoring standards
development organisations (SDOs), evaluating the evolution of
technical work, understanding technology deployment, and informing
community, leadership, and governance discussions.
This document is submitted for consideration by the Research and
Analysis of Standard-Setting Processes Research Group (RASPRG) in the
IRTF. It is not an IETF product and is not a standard.
About This Document
This note is to be removed before publishing as an RFC.
The latest revision of this draft can be found at
https://csperkins.github.io/draft-analysing-sdo-data/draft-perkins-
analysing-sdo-data.html. Status information for this document may be
found at https://datatracker.ietf.org/doc/draft-perkins-analysing-
sdo-data/.
Discussion of this document takes place on the RASPRG Research Group
mailing list (mailto:rasprg@irtf.org), which is archived at
https://mailarchive.ietf.org/arch/browse/rasprg/. Subscribe at
https://www.ietf.org/mailman/listinfo/rasprg/.
Source for this draft and an issue tracker can be found at
https://github.com/csperkins/draft-analysing-sdo-data.
Perkins, et al. Expires 7 November 2026 [Page 1]
Internet-Draft Analysing Internet Standards May 2026
Status of This Memo
This Internet-Draft is submitted in full conformance with the
provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF). Note that other groups may also distribute
working documents as Internet-Drafts. The list of current Internet-
Drafts is at https://datatracker.ietf.org/drafts/current/.
Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress."
This Internet-Draft will expire on 7 November 2026.
Copyright Notice
Copyright (c) 2026 IETF Trust and the persons identified as the
document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents (https://trustee.ietf.org/
license-info) in effect on the date of publication of this document.
Please review these documents carefully, as they describe your rights
and restrictions with respect to this document.
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Standards Development as a Socio-Technical System . . . . . . 4
3. Analysing the IETF . . . . . . . . . . . . . . . . . . . . . 7
3.1. Datatracker . . . . . . . . . . . . . . . . . . . . . . . 8
3.2. RFC Editor . . . . . . . . . . . . . . . . . . . . . . . 9
3.3. Mailing List Archives . . . . . . . . . . . . . . . . . . 9
3.4. Session Recordings . . . . . . . . . . . . . . . . . . . 10
3.5. Chat Archives . . . . . . . . . . . . . . . . . . . . . . 10
3.6. GitHub . . . . . . . . . . . . . . . . . . . . . . . . . 10
4. Analysing Other SDOs . . . . . . . . . . . . . . . . . . . . 10
4.1. Data Availability Across SDOs . . . . . . . . . . . . . . 11
4.2. Integrating Data Across SDOs . . . . . . . . . . . . . . 11
5. Data Processing . . . . . . . . . . . . . . . . . . . . . . . 12
6. Ethics and Data Protection . . . . . . . . . . . . . . . . . 13
7. Recommendations . . . . . . . . . . . . . . . . . . . . . . . 14
7.1. Recommendations for the IETF . . . . . . . . . . . . . . 16
7.2. Recommendations for Researchers . . . . . . . . . . . . . 17
8. Security Considerations . . . . . . . . . . . . . . . . . . . 18
Perkins, et al. Expires 7 November 2026 [Page 2]
Internet-Draft Analysing Internet Standards May 2026
9. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 18
10. Informative References . . . . . . . . . . . . . . . . . . . 18
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . 19
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 19
1. Introduction
Internet technologies are developed and standardised by a range of
standards development organisations (SDOs), including the IETF, W3C,
IEEE, 3GPP, ITU-T, and others. The standards that these
organisations produce underpin the interoperability and architectural
evolution of the Internet and the Web.
Understanding how standards are developed, including who participates
in the standards process, what collaborations occur during the
development of standards, how the process is organised and governed,
and the technical outputs, can support analysis of standards
ecosystems. Such analysis can assist with monitoring standards
development organisations, evaluating the evolution of technical
work, understanding technology deployment, and ultimately be used to
inform community leadership and governance discussions.
This document outlines considerations for studying data from the
Internet standards development ecosystem. It aims to:
* identify observable components of the Internet standards
development ecosystem;
* describe considerations for measuring and analysing the standards
development process;
* provide a taxonomy of possible measurements and analytical
approaches;
* highlight methodological, interpretive, and ethical
considerations;
* illustrate the application of these methods to the IETF, given the
availability of rich data about the IETF participants, documents,
processes, and communication channels;
* discuss the relevance and limits of applying these methods to
other SDOs and the extent to which differences in governance,
transparency, and data availability affect such analysis; and
* encourage reproducible research practises and transparent
analysis.
Perkins, et al. Expires 7 November 2026 [Page 3]
Internet-Draft Analysing Internet Standards May 2026
This document does not prescribe specific metrics, define evaluation
criteria, or recommend approaches to comparative rankings of
standards bodies, groups, or participants.
*TODO*: Cite prior work. This document currently does a very poor
job of citing prior work. This omission will be remedied in future
versions. While it is not intended that this document forms the
basis for a comprehensive literature survey, if you know of relevant
prior work that should be cited please contact the authors.
2. Standards Development as a Socio-Technical System
Internet standards development can be understood as a socio-technical
system in which technical artefacts, human participants,
organisational interests, and governance processes interact over
time. Standards do not emerge solely from technical design choices,
nor solely from institutional processes; rather, they arise through
structured collaboration among individuals and organisations
operating within formal and informal rules.
Technical outputs emerge from a socio-technical process in which
engineering choices interact with expertise, incentives,
organisational structures, review processes, historical precedent,
deployment constraints, and the cultural norms and practices of the
standards community. At the same time, the organisational and
cultural context is not fixed: governance structures, working
practices, and community norms evolve together over time and these
changes in turn shape future participation and technical decision-
making.
For analytical purposes, standards development ecosystems can be
viewed as comprising several interacting components:
* *Participants:* Participants are the individuals who contribute to
standards development. They may include engineers, researchers,
operators, implementers, academics, independent contributors,
civil society representatives, policy specialists, and others with
relevant expertise or interests. Participation criteria differ
across SDOs. Some use open participation, while others structure
participation through organisational- or state-based membership,
sometimes with additional exceptions or parallel open mechanisms.
Participation models affect standards development by shaping both
who is able to contribute, and how they are permitted to
contribute. Open participation can broaden the pool of
contributors and make it easier for individuals to join without
prior institutional affiliation, which may increase diversity of
experience and viewpoints. At the same time, openness does not
Perkins, et al. Expires 7 November 2026 [Page 4]
Internet-Draft Analysing Internet Standards May 2026
eliminate all the barriers to participation. Effective
participation may still depend on having sufficient time, funding,
employer support, travel resources, and familiarity with the
processes, tools, and norms of the community. Membership-based
models may provide clearer institutional commitment and
resourcing, but they can also limit participation to those acting
through recognised organisations or membership categories.
* *Organisations:* Participants are often affiliated with
organisations such as companies, academic institutions,
governments, consultancies, or civil society groups. These
organisations may provide forms of support including funding,
staff time, technical expertise, or implementation experience.
The relationship between participants and organisations is not
equally visible across SDOs. In some models, participation is
individual, and so any recorded affiliation may be incomplete, and
reflect a specific contribution rather than the sustained view of
the participant. In other models, where individuals participate
on behalf of a clearly indicated affiliation, the institutional
link is clearer.
Even where affiliations are recorded, they may not fully describe
the organisational context. A company may be a subsidiary of
another company (or in the process of becoming so), and
consultants or contractors may work for clients whose interests
are not directly visible in participation records.
* *Technical Groups:* SDOs typically organise work through technical
groups such as working groups, research groups, study group,
committees, or similar bodies. These groups define scope,
coordinate discussion, and develop technical outputs. They are
not always organised as a single flat layer, with hierarchical and
other structures in use.
The number, names, and functions of these structures differ across
organisations. In some cases, they reflect administrative
oversight or broad technical areas; in others, they distinguish
between different forms of technical development.
* *Artefacts:* Standards processes generate artefacts such as
drafts, specifications, recommendations, reports, agendas,
minutes, presentations, issue trackers, and final published
standards. These artefacts provide an observable record of
technical development. Revision histories, references, and
relationships between documents may help reveal aspects such as
participation dynamics, design iteration, and the evolution of the
underlying technologies subject to standardisation.
Perkins, et al. Expires 7 November 2026 [Page 5]
Internet-Draft Analysing Internet Standards May 2026
Different SDOs vary in how openly they make such information
available and in how easily it can be accessed and reused.
artefact availability can support the work of participants,
researchers, and other observers, but collecting, maintaining,
publishing, and organising this information also imposes costs on
SDOs.
* *Collaboration Infrastructure:* Standards development requires
communication among participants to propose work, discuss
technical issues, review contributions, coordinate activity,
resolve disagreements, and build support for possible outcomes.
SDOs therefore rely on systems such as mailing lists, code
repositories, and meetings to facilitate this debate.
The mix of communication, collaboration, and coordination
mechanisms differs across SDOs, often to support the other
attributes described.
* *Governance Structures:* Standards bodies have formal governance
structures, with charters specifying the scope of different
activities, defined leadership roles, review and approval stages,
appeals processes, voting rules, consensus procedures, and so on.
These structures define how work is initiated, scoped, reviewed,
approved, and contested.
At the same time, influence is also exercised through reputation,
recognised expertise, community norms, procedural familiarity, and
control over agendas, drafting, or review capacity. Governance
structures therefore shape how decisions are made, how priorities
are established, how disagreements are managed, and, ultimately,
how influence is distributed within standards development.
* *Standards Implementation and Deployment:* Implementation usually
occurs outside the formal standards process, and may be voluntary
by interested parties or mandated by policy in certain
jurisdictions.
In many cases, publication of a standard does not by itself
require implementation. Adoption may therefore vary widely: some
standards are widely deployed, while others see limited or no
implementation. Adoption may also be shaped by factors outside
the standards process, including regulation, procurement, cost,
and compatibility with existing systems.
Data on implementation and operational use is often limited.
Perkins, et al. Expires 7 November 2026 [Page 6]
Internet-Draft Analysing Internet Standards May 2026
Measuring SDO activity is challenging. Observable metrics such as
publication counts, message volume, attendance figures, authorship,
or leadership roles can provide useful evidence, but each captures
only part of the standards process. Analysis of artefacts and logs
from the collaboration infrastructure (e.g., analysis of mailing list
messages) can provide more detail and nuance, at the expense of
additional complexity, but even these do not provide a complete view.
There are several reasons for this. One is that critical aspects of
standards development are hard to observe directly. Influence,
agenda setting, informal coordination, negotiation, and the practical
exercise of power and authority may not be well represented by any
single metric, or group of metrics, and are extremely challenging to
infer from collaboration infrastructure logs.
Another reason is that the available data is often limited. Data
availability and quality vary across SDOs. Different parts of the
process are not equally observable, and even within a single SDO some
information may be incomplete, difficult to access, inconsistently
structured, or unavailable.
Combining multiple data sources introduces additional challenges.
Observations from different parts of the process may not share stable
identifiers, identifiers may change over time, and the same entity
may appear in different forms across records. Voluntary
declarations, non-standard terminology, and organisational changes
such as mergers or acquisitions may further complicate linkage.
Metrics, artefacts, and other data sources may also differ in
accuracy, representativeness, and relevance. Not all artefacts have
the same significance, not all forms of participation have the same
effect, and visible activity does not necessarily correspond to
implementation, adoption, or wider impact. Measures should therefore
be interpreted cautiously and, where possible, considered alongside
complementary indicators.
3. Analysing the IETF
IETF participation is open with no formal membership. Individuals
can participate by joining mailing lists, contributing to
discussions, submitting Internet-Drafts, and attending meetings.
Contributions ordinarily reflect the opinion of individual
participants, and not necessarily their affiliation; exceptions to
this norm exist for specific aspects such as draft authorship and
intellectual property rights disclosures.
Perkins, et al. Expires 7 November 2026 [Page 7]
Internet-Draft Analysing Internet Standards May 2026
The IETF has a hierarchical group structure, with technical working
groups (that have working group chairs) organised into distinct areas
(that have area directors).
Reflecting its open participation model, much of the IETF's processes
are publicly observable through open records and dedicated APIs.
Mailing lists are a central forum for working group discussion,
alongside meetings; some groups also use externally hosted
repositories, for example on GitHub, to support drafting and issue
discussion.
3.1. Datatracker
The IETF Datatracker (https://datatracker.ietf.org/) is the main
source of day-to-day and historical data about the operation of the
IETF. It can be accessed via the website or programmatically using a
REST API and provides information about:
* Participants including names, email addresses, pronouns,
biography, and photo, and external resources such as personal
websites, GitHub usernames, and Orcid identifiers. The
Datatracker maintains a record of the different names and email
addresses used by individuals.
* Artefacts such as RFCs, Internet-drafts, agendas, blue sheets,
working group charters, conflict reviews, shepherd write-ups,
liaison statements, minutes, and presentation slides, including:
- Metadata such as the title, name ("draft-ietf-..."), revision,
date, state, and where appropriate abstract, working group, RFC
number and publication stream, status on the standards track,
area director, and document shepherd.
- Submissions (e.g., different revisions of internet-drafts) with
document name, revision, date, title, abstract, authors, group,
and metadata about documents the submission replaces.
- Authors with email address, affiliation, and country.
- Events such as state changes state, expiration, details of IESG
processing, IETF last call, directorate reviews, IANA reviews,
etc., with the document name, revision, date, and responsible
person.
- Relationships including normative and informative references,
and document replaced, updated, or obsoleted.
Perkins, et al. Expires 7 November 2026 [Page 8]
Internet-Draft Analysing Internet Standards May 2026
* Working groups, research groups, area, directorates, and
leadership bodies such as the IESG, IRSG, and IAB, including the
group name and acronym, group state, relationships between groups
(e.g., working groups are organised in areas), the mailing list,
charter text, milestones, and who occupies key roles in the group.
* IESG processing, including ballot positions, the text of comments
and discusses, and scheduling of the IESG review.
* Directorate membership and directorate reviews, including the
document, reviewer, outcome, data, and the review text.
* Meetings, including both plenary and interim meetings, with
venues, dates, and times, details of what groups met in what time
slots, and registration and attendance data.
* IPR disclosures including the document that the IPR relates to,
the person making disclosure, details of the patent, and licensing
terms.
The Datatracker has been developed over time, and this is reflected
in the data that is available, with more recent data being
significantly more complete than earlier data. Datatracker profiles
are only required for a subset of IETF activities (e.g., draft
submission, meeting registration), and so a number of active
participants do not have a profile.
3.2. RFC Editor
The RFC Editor makes the RFC index available in machine readable form
at https://www.rfc-editor.org/rfc-index.xml. The RFC index includes
title, authors, publication date, status, abstract, publication
stream, name of the precursor Internet-Draft, and the IETF area and
working group that developed the RFC, if appropriate. This
information is also available in the IETF Datatracker.
Information about RFC errata is available on the RFC Editor website
at https://www.rfc-editor.org/errata.php. This data is also
available in machine readable form.
3.3. Mailing List Archives
The IETF maintains public mail archives at
https://mailarchive.ietf.org/ that are also available in machine
readable form via IMAP from imap.ietf.org. The recent mail archives
are essentially complete, but some historical lists that were not
originally hosted on ietf.org are missing. Spam emails have largely,
but not entirely, been removed from the archive. As of March 2026,
Perkins, et al. Expires 7 November 2026 [Page 9]
Internet-Draft Analysing Internet Standards May 2026
the IETF mail archive contains approximately 3 million messages from
almost 1400 mailing lists, around 40GB of data, with some messages
dating back to the late 1980s.
The are significant data quality problems with older messages in the
IETF mail archive, due to problems with the original messages rather
than the archive, that make them difficult to process.
3.4. Session Recordings
The IETF makes video recordings of its plenary meetings available on
YouTube (https://www.youtube.com/user/ietf). Audio recordings of
IETF sessions from IETF 49 through to IETF 106 are available at
https://get.ietf.org/archive/audio.
3.5. Chat Archives
The IETF makes chat logs available. Jabber was used prior to 2021,
with archives at https://get.ietf.org/archive/jabber/. More recently,
Zulip has been used accessible at zulip.ietf.org.
3.6. GitHub
Some IETF working groups, and some individuals, make extensive use of
GitHub for document development and issue tracking. The IETF does
not maintain a complete list of GitHub repositories associated with
its work. The IETF Datatracker contains links to some repositories
and user profiles.
Using the GitHub API, the following information is available:
* Information about GitHub users that contribute (e.g., username,
email address, and other biography information).
* Contributions and changes, by way of Git commits, made by those
users to documents.
* Discussion that takes place through comments and issues.
4. Analysing Other SDOs
Standards relevant to the Internet and the Web are also developed
within the W3C, 3GPP, ITU-T, and others. Each organisation has its
own governance model, participation structure, institutional culture,
and data availability. These differences affect both what can be
observed, and how observations should be interpreted.
Perkins, et al. Expires 7 November 2026 [Page 10]
Internet-Draft Analysing Internet Standards May 2026
4.1. Data Availability Across SDOs
SDOs vary considerably in terms of the data that they make publicly
available about their activities, and in how easily that data can be
accessed and processed.
The W3C provides a REST API at https://api.w3.org, covering metadata
about documents, participants, affiliations, and groups, and
maintains a public mailing list archive. W3C groups make extensive
use of GitHub for document development and issue tracking. The W3C
operates under a membership model, in which participation is
primarily through affiliated organisations. This affects how data
about participants and their contributions should be interpreted,
particularly when being compared to data from the IETF and other SDOs
with individual participation models.
The ITU-T and 3GPP both operate under membership-based models where
access to documents, meeting records, and contribution data is
typically restricted to member organisations. Some ITU-T
Recommendations are made publicly available after publication, while
the 3GPP makes its specifications available at https://www.3gpp.org/
specifications. The working documents, contributions, and meeting
records are generally not accessible to non-members.
Differences in data availability mean that the methods applicable to
the IETF, where rich longitudinal data is publicly available, may not
be replicable across all SDOs. Any analyses should account for these
availability differences.
4.2. Integrating Data Across SDOs
Efforts to understand the wider standardisation landscape requires
combining data across multiple SDOs.
SDOs do not share common identifiers for participants, organisations,
documents, or other metadata. An individual that participates across
multiple SDOs may appear under different names, e-mail addresses, or
usernames in the records of each SDO. Resolving these identifies
requires suitable entity resolution mechanisms, and the risk of both
incorrect matches (where two unrelated entities are linked together)
and missed matches (where one entity has multiple, separate records
in each SDO). The same risks apply to affiliations: companies may be
recorded under different names, abbreviations, or subsidiary
identities across SDOs.
Standards developed within one organisation may reference, build
upon, or be coordinated with work at another SDO, but these
relationships are not captured in any shared record. Reconstructing
Perkins, et al. Expires 7 November 2026 [Page 11]
Internet-Draft Analysing Internet Standards May 2026
these relationships requires either manual effort, or natural
language processing of document content, introducing the risk of
errors.
SDOs operate on different timescales and with different process
structures. Comparing activity across organisations at a given point
in time may not reflect equivalent stages of development.
Finally, differences in governance and participation models affect
which comparisons are meaningful. Data analyses, and the
interpretation of them, must consider that apparent differences
between SDOs may reflect structural factors (e.g., open vs.
membership-based participation) rather than substantive differences
in behaviour or outcomes.
5. Data Processing
Significant processing effort is required to clean, normalise, and
link data records before they can be processed.
The same individual participant may appear across each of the data
sources with different identifiers, including names, e-mail
addresses, usernames. These identifiers may change over time.
Entity resolution (using exact and heuristic matching) is feasible in
many instances, but requires careful validation to prevent the
introduction of errors into later analyses. Entity resolution across
organisations is similarly challenging, where companies may be
subsidiaries of another, might merge or be acquired, or, given the
unstructured nature of the dataset, appear under different names (to
illustrate the scope of the entity resolution problem note that, as
of May 2026, there are 282 variants of the name "Huawei" in the IETF
Datatracker). Information external to the Datatracker, and other
data sources, is often needed to process organisational data.
Participants may have more than one affiliation, including across the
lifetime of a particular contribution (e.g., an Internet-Draft).
Affiliation data is only recorded for a subset of activities, and may
need to be inferred (e.g., from corporate domain names) in other
cases. As a result, affiliation data, where recorded, indicates the
participant's affiliation at moment in time for a particular
contribution, making it difficult to form a continuous history.
Document life cycles are non-linear, and documents might pass through
multiple working groups, by replaced or updated by later drafts, and
change authorship over time.
Perkins, et al. Expires 7 November 2026 [Page 12]
Internet-Draft Analysing Internet Standards May 2026
Working group leadership is difficult to reconstruct: knowing who
chaired a working group during a particular period, or which area a
given group belonged to at a given time, requires the reconstruction
of a timeline from historical event records held in the Datatracker.
These records can be incomplete or inconsistently formatted.
E-mail metadata and message content presents a number of challenges.
A significant number of messages contain malformed or archaic header
fields that break widely used email processing libraries and need
correction. Mail clients perform the threading of messages in
different ways, with the separation between new and quoted text
becoming unclear. Natural language processing of message content
requires contextualisation, with informal conventions, technical
vocabulary, and the use of acronyms (all of which may evolve over
time) presenting challenges that are unique to the dataset.
As noted, the quality of the dataset degrades significantly for
historical records. Data that was not gathered by the Datatracker at
the time, or that has been subject to partial backfilling later, must
be treated with caution, both in terms of data processing and later
analyses.
6. Ethics and Data Protection
Data is made available by the IETF, and other Internet SDOs, subject
to their particular privacy and data protection policies and terms of
use. For the IETF, these are described at https://www.ietf.org/
privacy-statement/; other SDOs will have their own policies.
The available data includes considerable amounts of personal data
that is potentially sensitive and subject to legal restrictions on
processing and use in many jurisdictions (e.g., the GDPR in Europe).
Researchers must ensure that their use of such data conforms to any
applicable regulations. It is important to note that the regulations
that apply to research use of such data may differ from those that
apply to the IETF, or other SDOs, with regards to their use of the
data as part of the standards process.
Perkins, et al. Expires 7 November 2026 [Page 13]
Internet-Draft Analysing Internet Standards May 2026
Researchers must ensure that their research, in particular research
that involves personal data from the IETF or other SDOs, is conducted
ethically and with respect for persons, in careful consideration of
the risks and benefits of the work, taking care to ensure that those
who bear the risk also gain some benefit, and with respect for the
law and public interest. Researchers should consult with their
organisation's Institutional Review Board, Research Ethics Committee,
or similar, prior to conducting research that might raise ethical
concerns, and are referred to the guidance in the Menlo Report
[MENLO], the Belmont Report [BELMONT], and the ACM Policy on Research
Involving Human Participants and Subjects [ACM] for further
discussion of issues around ethical conduct of research.
Researchers are reminded that while data may be public, the
implications of that data are not always well-known. For example,
data that can be collected from the IETF Datatracker makes it
possible to derive measures of the effectiveness of individuals in
certain roles that, if presented out of context, might be considered
sensitive. It is inappropriate to publish data about specific
individuals without their explicit consent.
Finally, we note that researchers must take care to avoid disruption
to the Internet standards process. In part, this requires that they
consult with the operations staff in the IETF LLC, or other SDOs, to
ensure their data access does not cause operational difficulties
(e.g., overload of servers that might disrupt an ongoing meeting).
More broadly, researchers should ensure that any results that might
be considered sensitive or disruptive are responsibly disclosed to
the affected parties prior to publication. The effective operation
of the Internet standards process directly affects critical global
infrastructure, and researchers should be mindful of this when
presenting results.
7. Recommendations
Analysis of standards development data is useful to support
transparency and provide insight into the health, structure, and
evolution of standards ecosystems, including patterns of
participation, collaboration, concentration, and the development of
technologies. It can inform discussions within SDOs and provide
indicators of how technical work progresses over time. It can also
inform broader Internet governance questions, such as how decision-
making is structured, how participation is distributed, and the
extent of centralisation in these processes [RFC9518], and can be
useful to external stakeholders, including regulators, policy makers,
and civil society, seeking to understand how standards are developed
and governed.
Perkins, et al. Expires 7 November 2026 [Page 14]
Internet-Draft Analysing Internet Standards May 2026
Analysis of standards development is constrained by what can be
observed. Important aspects of the process such as informal
discussion, trust, institutional memory, cultural norms, and the
exercise of influence may be only partially visible. In addition,
the available data is often incomplete, inconsistently structured,
and shaped by changes in tools and processes over time, with
historical records in particular being sparse or unreliable.
As a result, analyses based on these data provide only a partial view
of the process. Quantitative metrics such as message volume,
authorship, participation counts, or leadership roles can be useful
indicators, but do not directly capture influence, authority, or
impact. They should therefore be interpreted with care and in
context, rather than in isolation.
Where data is derived or reconstructed (e.g., via entity resolution,
affiliation inference, or automated extraction) it is important to
retain a clear link to the original sources. The provenance of such
transformations should be documented, and derived data should be
distinguishable from primary records. This allows results to be
checked and, where necessary, corrected.
SDOs can support analysis of their processes by ensuring that the
data they produce remains consistent, well-structured, and accessible
over time. This includes maintaining clear, timestamped
documentation of artefacts and processes, recording changes and their
implications, and using consistent data formats and identifiers.
Providing structured access to data, for example through stable and
well-documented APIs can be especially helpful. When introducing
changes to tools, processes, or working practises, it is important to
consider how these affect what is recorded and how it can be
analysed. Where changes introduce discontinuities these should be
clearly documented, including their scope and implications, so that
their impact on the data can be understood and accounted for in
subsequent analysis.
Comparisons across standards development organisations require
particular care. Differences in governance, participation models,
and transparency affect both what is observable and how it should be
interpreted. Apparent differences between organisations may reflect
these structural factors rather than substantive differences in
behaviour or outcomes.
Perkins, et al. Expires 7 November 2026 [Page 15]
Internet-Draft Analysing Internet Standards May 2026
Finally, although much of the data used in this type of analysis is
publicly available, its use still raises ethical questions. Analyses
can have implications for individuals and organisations, especially
if results are presented without sufficient context. Researchers
should take care in how findings are reported, particularly where
they relate to identifiable participants.
7.1. Recommendations for the IETF
* *Preserving a centralised and stable data access:* The Datatracker
provides a central interface for structured data about IETF
activity. Maintaining this role, including stable identifiers,
consistent schemas, and well-documented APIs, supports
reproducible and longitudinal analysis. Where data is maintained
across multiple systems, stable references to authoritative
sources help ensure consistency and integration.
* *Data quality and consistency:* The data reflects changes in tools
and practices over time, which can make it harder to interpret,
especially for older records. Common data such as events, roles,
group metadata, and document states may be inconsistent across
time. Where possible, these differences should be made consistent
or clearly documented.
* *Historical data and backfilling:* Historical data may be
incomplete. Where records can be reconstructed with confidence,
backfilling can improve coverage. Backfilled data should be
clearly identified, and its provenance documented.
* *Provenance of derived data:* Where data is derived from primary
sources (e.g., extraction from archival material), the
relationship between source and derived data should be explicit.
Original artefacts should be retained where possible, and derived
records clearly distinguished to allow validation and correction.
* *Error reporting and correction:* Datasets will contain errors,
particularly in historical or reconstructed records. Providing a
transparent mechanism for reporting and correcting errors, along
with maintaining a record of changes, improves reliability.
* *Separation of primary and inferred data:* Some data useful for
analysis (e.g., identity resolution, affiliation inference)
involves interpretation. Such data should be distinguishable from
primary records, with clear documentation of how it was produced.
- *TODO:* is this done by the IETF, by the researchers, or both?
Perkins, et al. Expires 7 November 2026 [Page 16]
Internet-Draft Analysing Internet Standards May 2026
* *Impact of process and tooling changes:* Changes to tools and
working practises affect what is recorded and how it can be
analysed. Where such changes introduce differences in data
structure or coverage (e.g., adoption of different collaboration
platforms), these should be documented clearly, including their
scope and implications, to preserve comparability across groups
and over time.
7.2. Recommendations for Researchers
Analysis of standards development data requires careful handling of
both the data and its interpretation. The following practises can
improve the robustness and reproducibility of such work:
* *Care in Datatracker use:* When using the Datatracker, it is
preferable to download a local snapshot of the data, while
respecting any access limits, and perform analysis on that copy.
This avoids repeated queries to the live API.
* *Use versioned data snapshots:* The underlying datasets evolve
over time. Analyses should be based on well-defined snapshots
rather than live data, so that results can be reproduced and
compared.
* *Document data processing steps:* Significant processing is often
required before analysis, including cleaning, normalisation, and
entity resolution. These steps can materially affect results and
should be clearly documented, including any assumptions or
heuristics used.
* *Handle identity and affiliation data with care:* Participants may
appear under multiple identifiers, and affiliations may be
incomplete, ambiguous, or change over time. Methods used to
resolve identities or infer affiliations should be validated where
possible and treated as approximations.
* *Account for incomplete and inconsistent data:* Not all aspects of
the standards process are equally observable, and available data
may be incomplete or inconsistent, particularly for historical
records. Analyses should account for these limitations and avoid
over-interpreting gaps or trends.
* *Be cautious in interpreting metrics:* Common metrics such as
message volume, authorship, or participation counts do not
directly capture influence, authority, or impact. Results should
be interpreted in context and, where possible, supported by
complementary evidence.
Perkins, et al. Expires 7 November 2026 [Page 17]
Internet-Draft Analysing Internet Standards May 2026
* *Consider the impact of tooling and process changes:* Changes in
tools or working practises (e.g., use of different collaboration
platforms) can affect what is recorded and how it is structured.
These changes should be considered when interpreting longitudinal
trends or comparing across groups.
* *Engage with the community:* Data alone provides an incomplete
view of the standards process. Engagement with participants or
domain experts can help interpret results and identify factors
that are not visible in the data.
* *Support reproducibility and reuse:* Where possible, researchers
should share datasets, code, and methods, subject to applicable
policies and privacy considerations. This reduces duplication of
effort and improves the reliability of results.
* *Contribute improvements where appropriate:* Effort spent cleaning
or structuring data may be of broader value. Where feasible,
contributing corrections or improvements back to shared data
sources can benefit the wider community.
* *Consider ethical implications:* As discussed in the Ethics and
Data Protection section, analysis may have implications for
individuals or organisations. Care should be taken in how results
are presented, particularly where they may be sensitive or open to
misinterpretation.
8. Security Considerations
Research into the operation of the Internet standards development
ecosystem does not directly affect the security of the Internet.
Effective operation of the Internet standards process is, however,
critical to the security of the network, and researchers studying the
development of Internet standards must consider potential security
implications of their results and ensure that any such implications
are responsibly disclosed to the relevant SDO. Examples might
include, but are not limited to, research that discovers attempts to
subvert or disrupt the operation of the standards process.
9. IANA Considerations
This document has no IANA actions.
10. Informative References
Perkins, et al. Expires 7 November 2026 [Page 18]
Internet-Draft Analysing Internet Standards May 2026
[ACM] ACM Publications Board, "ACM Publications Policy on
Research Involving Human Participants and Subjects", n.d.,
<https://www.acm.org/publications/policies/research-
involving-human-participants-and-subjects>.
[BELMONT] National Commission for the Protection of Human Subjects
of Biomedical and Behavioral Research, "The Belmont Report
- Ethical Principles and Guidelines for the Protection of
Human Subjects of Research", n.d.,
<https://www.hhs.gov/ohrp/regulations-and-policy/belmont-
report/>.
[MENLO] US Department of Homeland Security Science and Technology
Directorate, "The Menlo Report - Ethical Principles
Guiding Information and Communication Technology
Research", August 2012,
<https://www.dhs.gov/sites/default/files/publications/CSD-
MenloPrinciplesCORE-20120803_1.pdf>.
[RFC9518] M. Nottingham, "Centralization, Decentralization, and
Internet Standards", December 2023,
<https://datatracker.ietf.org/doc/html/rfc9518>.
Acknowledgments
This document builds on work funded, in part, by the UK Engineering
and Physical Sciences Research Council under grants EP/S033564/1 and
EP/S036075/1.
Authors' Addresses
Colin Perkins
University of Glasgow
Email: csp@csperkins.org
Ignacio Castro
Queen Mary University of London
Email: i.castro@qmul.ac.uk
Ryo Yanagida
University of St Andrews
Email: ryo@htonl.net
Stephen McQuistin
University of St Andrews
Perkins, et al. Expires 7 November 2026 [Page 19]
Internet-Draft Analysing Internet Standards May 2026
Email: sjm55@st-andrews.ac.uk
Perkins, et al. Expires 7 November 2026 [Page 20]