16 What the record does not tell you

16.1 Why a chapter of holes is worth reading

Everything in this course comes from two places: the clauses of TS 38.300, and a register built from files and databases on one machine on 4 August 2026.

Both are good sources and neither is complete. A course that only says what it knows leaves the reader to assume the rest is unknowable or, worse, to assume it is covered. This chapter says exactly where the edges are.

16.2 G1. No person is named, anywhere

The register carries the organisation or the role a document came from — "RAN Chair (Samsung)", "ETSI MCC", "Nokia" — and never a human being's name RP-260830 RP-260831.

So this course cannot name a chair, a vice chair, a rapporteur, a delegate or an author. Not for any group and not for any document.

Where to look: each group's chair history is its own page on 3gpp.org, addressed by the group's code ^[1].

16.3 G2. No reason for any decision

A document's status says approved, revised, noted, withdrawn or not pursued, and stops there. RP-261197 is recorded as "not pursued" and the record does not say why. R1-2603888 is "not treated" and does not say why either.

This course therefore never writes why something was agreed or rejected. Where a chapter says a thing happened, it means the record shows it happened.

Where to look: the meeting report — RP-260831 is the plenary's — and the discussion papers of the meeting, found through the document list ^[2].

16.4 G3. No versions, no release dates, no freeze dates

The register holds no version of any specification, and no date on which a release opened or closed. It can tell you that 195 pieces of work belong to Release 19 ^[3] and cannot tell you when Release 19 started or ended.

The one version number anywhere in this course is inside a document title: RP-260870 is "TR 38.914 v0.4.1", and that is the title of a paper, not a statement about the document's current version.

Where to look: a specification's information page carries its versions and releases ^[4], and the file folder holds every published version ^[5].

16.5 G4. Almost no document text

The register holds titles and addresses. 470 of the 1059 active RAN documents are parsed into readable text on this machine ^[6]. The rest are not, and none of RAN6's are.

That is why this course quotes clauses of TS 38.300 and never of any other RAN document. Where another specification is named, it is named as a document — never as a sentence, and never with a clause number.

Where to look: the file folder of the document ^[5].

16.6 G5. No comparison with the rest of 3GPP

The register was built from RAN's rows only. Two totals for all of 3GPP exist — 3599 active documents ^[7] and 1629581 meeting documents ^[8] — and they are there to size RAN's share, not to compare groups.

Nothing here measures SA, CT or any other group, so this course says nothing about whether RAN is larger, faster or busier than any of them.

Where to look: the same status report the whole catalogue comes from ^[9], which covers every group.

16.7 G6. No membership, no company counts, no votes

There is no list of member organisations, no count of companies attending, and no record of how anything was voted on.

Attendance is the only people-number in this course, and only for meetings the portal recorded it for: 710 at RAN1#125 ^[10], 396 at RAN#112 ^[11] and the four others in What the numbers say about the size and shape of each group.

Every one of those is the number of people in the room.

Where a meeting was a teleconference, the record has nobody in a room at all. RAN6's 3GPPRAN6#18-Decision Teleconference leaves the field out rather than printing a zero, which is the honest treatment and also means the group's real participation is unknown here.

Where to look: the meeting's portal page ^[12].

16.8 G7. No full meeting archive on this machine

Only 23 RAN meeting folders are mirrored here, the most recent ones ^[13]. The meeting database reaches much further back than the files do.

So a meeting can be dated, counted and cited in this course while none of its documents are readable locally. Never say a document "is on this machine" unless you have opened it.

Where to look: the document's own address on the file server ^[14].

16.9 G8. Every count is a count of this machine

The counts here are counts of what the local catalogue and the local meeting database hold, on 4 August 2026. They are honest about the database and silent about the world.

Where the difference could matter, the chapters say "the local record holds". Read every number in this course that way, including the ones that do not say it.

Where to look: the 3GPP status report, for the catalogue ^[9], and a meeting's own document list, for the papers ^[2].

16.10 G9. The catalogue itself is not perfect

Three active rows carry no usable group at all ^[15]. That is a small number against 3599, and it means every per-group total in this course is a very good number rather than a proven one.

Where to look: count it yourself from the status report ^[9] if the total has to survive being checked.

16.11 G10. Almost nothing about 6G

One fact exists: a study report numbered 38.914 TR 38.914 and one draft of it agreed at RAN#112 RP-260870. No date, no scope, no content.

Anything else you have read about 6G in 3GPP did not come from here, and this course cannot confirm or contradict it.

Where to look: the study report's own information page ^[4] and the plenary's document list ^[2].

16.12 G11. "Study" is a database field

The word "study" throughout this course is a document type in a database, not a defined 3GPP term. When the difference carries weight in a sentence, say "a document the database records as a study".

The same caution applies to "work item": 3230 pieces of work were opened by a RAN plenary document ^[16], counted as folders whose metadata names such a document. That is a defensible definition and it is the one used here.

16.13 What this course does claim

Having listed the holes, it is worth saying what is solid.

Every clause quoted in the five TS 38.300 chapters is the document's own text, checked word for word when the site is built. Every number cited carries the question that produced it and the day it was asked.

Every document, meeting and group named exists in the register, and a name that did not would stop the site building at all.

That is a narrower promise than "this course tells you about TSG RAN". It is the promise worth making.

16.14 Where to go from here

The three depths of this track cover the same material at three altitudes; this is the detailed one.

The document behind it is TS 38.300, and every clause cited here opens in the panel beside the text. The register behind it is listed on this course's sources page, entry by entry, with the method for every count.

To go further into the radio itself, the four documents to read next are TS 38.211, TS 38.212, TS 38.213 and TS 38.214 for the physical layer, or TS 38.331 for the protocol.

To go further into how the group works, start at How to follow TSG RAN yourself, from a cold start and pick a meeting.