7 Reading either body's paper trail yourself

7.1 What a 3GPP number says before you open it

Two things are readable off the front of any 3GPP document.

The first pair of digits gives the subject area — requirements, service aspects, technical realization, signalling protocols and so on — under a table of number ranges §4.0. And TS against TR tells you whether you are holding a specification or a report.

7.2 What the version says

Three digits, and each one is a different kind of fact §4.0A, ^[2]:

• the first is the release once the document is under change control, and before that the state of the draft;
• the second goes up on every technical change;
• the third goes up on a purely editorial one, and resets when the second moves.

So 20.2.0 is a Release 20 document that has taken two rounds of technical change since it entered the release, and no editorial-only fix since the last of them.

7.3 What the file name says

The two facts are packed into the file name as well. The document's own example is the shortest explanation §5A:

Draft versions substitute a "d" for the hyphen, and those files never appear in the official directories §5A.

7.4 Where the history of a clause is written

Every 3GPP specification ends with a change history annex. Each row names the change request that changed the document, the meeting it was approved at, and the version that came out §Annex A. Editorial fixes made by the Support Team are explained there too §4.6.6.

That annex is usually the fastest way to answer "why does this clause say this". From it you have a change request number, and from there:

• the full change-request database, on the 3GPP file server ^[3];
• the portal, which shows the change requests of one specification or one meeting together with the decisions taken on them ^[4];
• the website, with the meeting calendars, the minutes, the documents and the latest specifications §7.6, ^[5];
• 3GPP's own step-by-step guidance on filling in the form, if you are about to write one ^[6];
• and this machine's own viewer, which shows the working-methods document clause by clause ^[7].

7.5 Everything at a meeting is a TDoc

One piece of vocabulary unlocks the meeting record.

The name is short for Temporary Document and is a leftover from paper, when such documents were not meant to survive the meeting §9.1. Nothing is thrown away now.

The permitted kinds are a fixed list — agendas, liaison statements in and out, pseudo change requests, draft change requests, change requests, change-request packs, work item descriptions, status reports, draft specifications, reports, discussion papers and a handful more §9.1. The statuses a document may end with are a fixed list too §9.2.

Counted across all 1,629,581 meeting documents on record ^[8], three of those kinds are worth having a feel for.

Kind of document	How many
Liaison statements sent to another group	47,474 [9]
New work item descriptions	11,595 [10]
New study item descriptions	5,532 [11]

7.6 Available is not published

A 3GPP document being downloadable does not make it a standard anywhere.

Approved versions go on a file server that allows anonymous access to any interested party §5.1. Formal publication is a separate act: the Organizational Partners that are standards bodies publish 3GPP's output as their own standards, under their own rules, which the working-methods document explicitly puts outside its own scope §5.1.

7.7 What "shall" and "MUST" mean

Both bodies define their requirement words once and reuse them everywhere, and both definitions live outside the document you are reading at the time.

In 3GPP, "shall" makes a requirement ^[12] and "should" makes a recommendation ^[13]. "Must" is kept out of the drafting rules altogether, so that nobody reads a standard as a law. Those wording rules are their own document TR 21.801 v19.1.0, and the shared vocabulary is another: TR 21.905 v19.2.0 defines 579 terms ^[14].

In the IETF, MUST is an absolute requirement ^[15], SHOULD allows a reasoned exception ^[16], and MAY is truly optional ^[17] — all three defined once, in RFC 2119.

An RFC's own header block is the other half of reading one: it gives the category, the BCP or STD number if it has one, and what the document obsoletes or updates RFC 7841.

7.8 What the record does not tell you

The paper trail is very good at some questions and silent on others. The silences here are worth knowing, because they are the ones people fill in by guessing.

• Dates before 2015. Of those 1,629,581 meeting documents, only 1,543,057 carry an upload date, and those run from 2015 to 2026 ^[18] — even though the meetings go back to 1998. The rest carry no readable upload date at all.

• Meetings that have not happened. The calendar records 4,371 3GPP meetings, the last of them planned for 2028-06-13 ^[19] — it holds scheduled meetings alongside held ones, so its last entry is a plan, not a record.

• Who anybody is. Nothing read here says how many delegates attend, how many companies are members, what membership costs, or who chairs anything.

• Freeze dates. The working-methods document says they are set in the work plan; the work plan is not among the files read here, so this course names no date for any release §4.10.3.4.

• The IETF's meetings. The archive holds RFCs and nothing else — no charters, no minutes, no mailing lists, no drafts ^[20].

Every number this course prints carries, in the register behind it, the exact query that produced it and the day it was asked. The entry behind the 1,629,581 meeting documents names the table, the database file and the date, and so does every other one. That is the point of the register: a figure whose origin nobody can restate is a figure nobody should repeat.

7.9 Where to go next

The working-methods document itself is the shortest route deeper, and it is readable clause by clause on this machine — TR 21.900 v20.0.0, ^[7]. The 3GPP website carries everything the file server does not ^[5], and the current IETF working groups and areas live only on the IETF's own site ^[20].

Every source this course cites, with its address and, for each number, the query behind it, is listed on the course's sources page.