The Basics
Reading RFCs
Writing RFCs
Authors’ Final Review (AUTH48 State)
The RFC Editor is no longer a single person; it is a small group of people. TheIETF Administration LLC (IETF LLC) contracts the RFC Editor function toAssociation Management Solutions, LLC (AMS). Through 2009, the home of the RFC Editor function was the Networking Division of theUSC Information Sciences Institute (ISI) in Marina del Rey, CA. ISI played a key role in the development of the Internet, and Jon Postel was the Director of ISI’S Networking Division for many years. For a historical account of the RFC series, see“30 Years of RFCs,”“40 Years of RFCs,”, and“Fifty Years of RFCs.”
This label in the heading of RFCs is historical in form and symbolic in content. Historically, “Network Working Group” meant the set of researchers who developed the packet switching protocols for the ARPAnet, beginning in 1969. This label was maintained in RFCs as a reminder of the long and significant technical history that is recorded in the RFC series, and as a reminder that today’s technical decisions, wise or not, may be with us for many years. Today, the “Network Working Group” should be interpreted as the set of users, vendors, and researchers who are working to improve and extend the Internet.
In a word,“NO!”. Many RFCs have Informational or Experimental status and do not represent any kind of standard. They contain information that may be useful or important to retain in this archival document series. This is important to understand, because unscrupulous marketeers and a careless trade press sometimes falsely suggest that every RFC represents a standard, or that all standards have equal weight. The relationship among Internet technical specifications is often complex.
Consult the online document“Internet Official Protocol Standards”.
Yes, the status of an RFC can change; this information is available in several locations including the RFC info page and RFC search results. For example, an RFC can be moved from Proposed Standard to Internet Standard (as described inRFC 6410) or from Informational to Historic. For a list of all RFCs that have changed status, please see thelist of status changes.
Youcannot! Once an RFC is published, it cannot be changed. The RFCs form an archival series. If the bug represents a change of content, a revised RFC can be written that obsoletes the one in error. For both technical and editorial errors, the RFC Editor provides a list of errata for published RFCs. Use theRFC Errata page to look up errata by RFC number or view the complete list. Also, search results from theRFC search page include hyperlinks to any corresponding errata entries. To report an error in an RFC, please use the form available from the RFC Errata page (seeHow to Report Errata for details).
For RFCs that have verified technical errata, there are files available with the errata shown inline (when possible). They are listed as “HTML with inline errata” in the RFC search results and info pages. For example:
RFC 5234.
All RFCs may be freely reproduced and translated (unmodified). Since the publication ofRFC 5377 andRFC 5378 in November 2008, the copyright notice and legends that appear in RFCs have been determined by theIETF Trust Legal Provisions. See theIETF Trust Copyright FAQ for further information.
Document Stream | Standards Track* | Best Current Practice (BCP) | Experimental | Informational | Historic |
---|---|---|---|---|---|
IETF | X | X | X | X | X |
IRTF | X | X | X | ||
IAB | X | X | X | X | |
Independent | X | X | X |
* including Internet Standard (STD) and Proposed Standard
Stream | Stream Manager |
---|---|
IAB | IAB Chair |
IETF | IESG |
Independent Submissions | ISE |
IRTF | IRSG |
Yes. An announcement of each new RFC is sent to all members of the rfc-dist mailing list. You can subscribe and unsubscribe from this list athttps://mailman.rfc-editor.org/mailman/listinfo/rfc-dist. There’s also anRSS feed andAtom feed.
There is alist of RFC numbers that were issued to documents that were never actually published. This explains the occasional gap between numbers. The current procedures are set up to try very hard to avoid this situation in the future. In particular, RFC numbers are never reserved; rather, they are assigned at the last moment in the editorial process.
In addition, some RFCs prior to 800 existed only on paper. The RFC Editor has an“RFC Online” project to make the entire RFC series available online. However, this process has necessarily had lower priority than editing new RFCs. We are grateful for the help of volunteers in the Internet community who entered and nroffed text of the missing online RFCs.
See“The End-of-Line Story” for a historical account of the problem and possible solutions.
The RFC Editor does not publish the repository in hard copy. There are several reasons for this. First, with over eight thousand RFCs the size of a hard copy would fill several bookcases. Second, given that most of the community obtains electronic versions of these documents, there is insufficient market to justify the printing costs. Finally, the RFC repository is constantly being extended. Any printed version would be quickly out of date.
Seethe RFC publication process.
Please look at theRFC Editor Queue.
Typical time to publish is 1-2 months, but the actual time varies greatly from one RFC to another. Publication may be held up for a variety of reasons, including IESG approval, inconsistencies or omissions that show up in editing, or normative references to other documents that must be published earlier or simultaneously. (For current information on documents linked by normative references, see thecluster page.) Authors should also be aware that the RFC Queue may be congested right before meetings of the IETF.
No, reserving an RFC number is not an option. The RFC number assigned to a given document is released when the document moves to AUTH48 state. For further information as to why this policy was adopted, please seeI cannot retrieve the text of an RFC. Why not?
If expedited publication is needed (including release of an RFC number only) to coordinate publication with other standards development organizations (SDOs), contact your stream manager. They may request that your document be prioritized. A number will be released when AUTH48 is initiated. Please be sure to notify the RPC of the date by which the RFC number is needed.
If your document is in theRFC Editor Queue, please go ahead and send the changes to the RFC Editor at any time.
See theRFC Style Guide. Also, we generally refer to“The Chicago Manual”.
References to RFCs should appear as described in theOnline Portion of the RFC Editor Style Guide; see “Referencing RFCs”.
We recommend using these files for referencing RFCs:
XML: viaBibXML Service (e.g.,https://bib.ietf.org/public/rfc/bibxml/reference.RFC.5234.xml)
TXT:https://www.rfc-editor.org/in-notes/rfc-ref.txt
Representative reference tags (such as [RFC2119]) are preferred over numeric reference tags (such as [1]).
References to I-Ds should appear as described in
Section 4.8.6.3 of the intended update to RFC Style Guide.
We recommend using the XML citation library; the URLs are of the form https://datatracker.ietf.org/doc/bibxml3/reference.I-D.foo.xml. For example:
<xi:include href=”https://datatracker.ietf.org/doc/bibxml3/draft-ietf-mmusic-msrp-usage-data-channel.xml”>
xml2rfc will reference the most current version of the I-D when this format is used. If a reference to a specific version of the I-D is desired, included the version number as part of the URL.
Yes, during AUTH48 state. SeePublication Process andAuthors’ Final Review (AUTH48).
We recommend one of the following paths:
Option 3 is typically used in instances where the missing author made significant contributions to the document, so the other authors are not comfortable removing the individual from the author list.
The WG chairs and Area Directors are CC’ed on every message sent from the RFC Editor to the authors during the course of the publication process, so they have the option of giving input at any time. Specifically, Area Director approval is requested when any changes that are beyond editorial are introduced into the document. A reply from WG chairs may be sought for issues that affect more than one document from their WG or for decisions about whether the WG needs to review changes.
The Document Shepherd (when not one of the WG chairs) is also CC’ed on each message from the RFC Editor during the publication process.
After the RFC is published, the authors as well as the WG chairs and Area Directors receive the notification message iferrata are reported for that RFC.
You can use SVG; please seeHow do I include SVG in my document?.
For background, before the transition to v3 XML, there was an option to post a PDF with enhanced images after the ASCII text was published. It would contain the exact text of the RFC with diagrams added. This file was produced by the authors. Since 16 September 2020, authors may no longer submit enhanced PDF files, as they can use SVG to create their figures in the RFC.
Note that the enhanced PDF files posted before that date are still available. Examples of RFCs that have used this option areRFC 4128,RFC 4137,RFC 4601,RFC 5059,RFC 5317, andRFC 5598. The optional PDF is listed as “PDF with images”.
For a recent source file (RFC 8650 and onwards), please retrieve it fromhttps://www.rfc-editor.org/in-notes/prerelease/. This is the“not prepped” XML source file, best for continued editing.
For numbers below RFC 8650, source files (XML or NROFF) are available by request; please send mail to theRFC Editor.
It’s the file before the prep tool defined inRFC 7998 is run. Before we publish the source file of an RFC, we use the xml2rfc tool to do a process called “preparing” the XML. This fills in defaults, adds explicit section numbers, and adds other data (including any external content) to make it easier to render the XML consistently into other formats. While the prepped XML is ideal for rendering, it is far from ideal for editing, since section numbers can change as text is added or reordered, and all of the default values can make it hard to find the useful XML elements and attributes.
April 1st submissions are the only RFCs-to-be that do not need to be posted as Internet-Drafts. These entries should be sent directly to theRFC Editor. We appreciate receiving all entries at least 2 weeks prior to April 1st so that the RFC Editor team has time to review all of the documents and prepare those that we decide to publish.
In the browser, there are several options to save the XML file. Please use “view page source” (or similar) and save the file.
where XXXX is the number of the RFC.
Note: If there are questions included within the XML file, you do not need to reply in both the XML file and the email. Replying in either form is sufficient. Please seethe AUTH48 notice for more information.
where XXXX is the assigned RFC number.
Note: Before or after your approval of the document as a whole, there may be changes sent by your coauthors. Assuming you have already sent us your approval of the document, it still stands unless you notify us otherwise. Any changes that are beyond editorial will be sent to the relevant body for approval.