School of Specs Inside one plenary: TSG SA#112In depth

What a plenary is · chapter 2 of 12 · 9 minutes

2 SA#112 as arithmetic

How many documents, of what kind, from whom, and ending in which word — and which two counts of the same thing disagree.

2.1 What a number from a meeting is worth

Every figure in this chapter is a count of paper. It says how many documents did something, never how much any one of them mattered.

That is worth saying once and remembering throughout. A meeting that approves two hundred documents has not necessarily done more than one that approves twenty, and nothing in this record lets you tell.

What the counts are good for is shape. They tell you where the meeting spent its four days, which release it was really working on, and which questions were settled and which were only read out.

2.2 The meeting on one page

What How many
Documents 346 [1]
Kinds of document 20 [2]
Different sources submitting 62 [3]
Top-level agenda items 23 [4]
Individual change requests 689 [5]
Specifications changed 114 [6]
Registered 568 [7]
Checked in 365 [8]
Registered to attend online 110 [9]

Four days, 346 documents. That is the whole scale of the thing, and it is why nobody reads a plenary front to back.

2.3 By kind of document

The register breaks out ten of the twenty kinds. Largest first:

Kind How many
change-request pack 120 [10]
new work item description 53 [11]
report 29 [12]
draft technical report 28 [13]
discussion paper 20 [14]
incoming liaison statement 18 [15]
standalone change request 14 [16]
other 13 [17]
new study item description 10 [18]
revised study item description 9 [19]

Those ten kinds carry almost every document of the meeting. The other kinds share what is left between them, and the register does not break them out.

Two things stand out. Change-request packs are the largest kind by a wide margin — a plenary is largely a machine for signing off corrections. And new work item descriptions are the second biggest kind, which is the other half of what a plenary is for: deciding what may be written next.

2.4 By what happened to it

Ten words cover every document of the meeting, and they add up to exactly 346:

Status Documents
approved 199 [20]
noted 68 [21]
revised 60 [22]
postponed 5 [23]
endorsed 4 [24]
partially approved 4 [25]
merged 2 [26]
withdrawn 2 [27]
agreed 1 [28]
replied to 1 [29]

Those words are not interchangeable and chapter The kinds of document, and the ten words that decide them takes each one apart. The headline is that "approved" is much the largest of the ten, and "noted" — read and filed, no decision taken — is the next.

"Revised" is the third biggest, and it is the one that surprises people: sixty documents at this meeting ended by being replaced with a different document number. That is not failure. It is the normal way a paper gets fixed during the four days it is in the room.

2.5 By release

Every document either belongs to a release or does not:

Release Documents
none 162 [30]
Rel-20 140 [31]
Rel-19 26 [32]
Rel-21 11 [33]
Rel-18 5 [34]
Rel-17 2 [35]

These also add up to exactly 346, which is a useful check that you are looking at one question asked of one set of rows.

The largest group carries no release at all — agendas, reports, liaison statements, project management. The second largest is Release 20, the release being written, and chapter Release 20 at this meeting, and what the numbers will not say is about what that means.

2.6 By who sent it

62 different sources submitted something. The five busiest are all working groups.

Sender Documents
SA WG5 66 [36]
SA WG3 57 [37]
SA WG6 41 [38]
SA WG2 33 [39]
SA WG4 28 [40]

No single company comes close to any of those five. That is not a claim about who was influential — it is a claim about who submits, and a working group's submission is often the collected work of many companies.

What each group works on can be read off the specifications its change requests touch, and nothing here says more than that. SA WG5's changes land on TS 28.552 and TS 28.312; SA WG2's on TS 23.501; SA WG6's on TS 23.482.

That is a deliberately thin claim, and it is the only one this data supports. A working group's remit is written in its terms of reference — and at this meeting every group's terms of reference was postponed [41], so even 3GPP had not settled the question in June 2026.

Every other source is the tail: smaller working groups, individual companies, groups of companies signing one paper, and 3GPP's own support staff. One document at this meeting was signed by ten public bodies at once SP-260584, and the database counts that whole list as a single source.

2.7 Where these numbers come from

Every count in this course carries the question that produced it and the day it was asked. Follow any of the links above and you get the exact query — which table, which column, which filter — rather than a figure somebody typed.

That is not decoration. A number about a meeting is only as good as the question behind it, and the same meeting will give you different numbers for "how many change requests" depending on what you meant. The rest of this chapter is that problem in miniature.

Three checks are worth doing on any meeting's figures before you quote them.

  • C1. Do the parts add to the whole? The ten status words add to 346. The six release groups add to 346. Both times the answer matches the meeting's own document total, so both questions were asked of one set of rows.

  • C2. Do two sources agree? The report's own annex closes with 346 entries [42], counted at the meeting, and the database says the same. So does the packs annex, 120, against the database's 120.

  • C3. Does a subtotal cover the total? The register breaks out ten kinds of document, but 20 different kinds turned up, so any sentence about "all the document types" is about ten of those 20 and should say so.

2.8 Where two counts of one thing disagree

The change requests of this meeting can be counted two ways, and the two answers differ:

  • 689 [5] — every token in the specification column of every change request and every pack.

  • The four release counts added together — 487 for Rel-20 [43], 143 for Rel-19 [44], 36 for Rel-18 [45] and 7 for Rel-17 [46] — which come to less.

Both are right. They are answers to two different questions: the first looks at every change request and pack, the second only inside packs.

2.9 The numbers that count the report, not the meeting

Some of these figures come from the secretary's report rather than from the document database, and the report is a draft [47]. Its annex tables each close with a count of their own entries:

  • 346 documents — the same total, arrived at independently.

  • 120 change-request packs — again the same figure the database gives.

  • 26 specifications and reports approved [48] and 6 noted for information [49].

  • 30 new work and study items approved [50] and 11 updated [51].

  • 18 letters in [52] and 2 out [53].

  • No work item exception sheets at all [54] — that annex reads "No entries found."

Two of those agree with the database exactly, which is the best evidence you will get that the paperwork of this meeting hangs together.

Two more counts describe the record rather than the decisions. Only 284 of the 346 documents get a paragraph of their own in the report [55]; the rest appear only as a row in a table. And the sentence "There were no contributions to this agenda item." appears 40 times [56], which is the sound of an agenda built for every possible meeting being used by one actual meeting.

Outside 3GPP's own files there is one more: 224 of the meeting's documents have a machine-written summary [57]. Those are written by a language model, they are useful for finding your way around, and chapter What this record does not tell you says plainly why they are never evidence of what was decided.

2.10 Where this leads

You now know the size and shape of the meeting. The next question is how it was organised, which is the agenda: How to read a plenary agenda.

Where the numbers in this chapter come from

  1. 346 documents at SA#112 sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112';", read 2026-08-04
  2. 20 different kinds of document distinct values of tdoctype over the same 346 rows, read 2026-08-04
  3. 62 different sources submitted documents distinct values of source over the same 346 rows. A source is a working group or a company, and one entry may name several, read 2026-08-04
  4. 23 top-level agenda items rows of the agenda table in SP-260361_SA112_Agenda_v2.md whose item number has no dot in it, read 2026-08-04
  5. 689 individual change requests the crspec column of every CR and CR pack row of tdoc for meeting='SP-112', split on spaces and counted. Each token is one change request against one specification, read 2026-08-04
  6. 114 specifications were changed distinct tokens of the crspec column over the same CR and CR pack rows, read 2026-08-04
  7. 568 people registered for the meeting the '568 Entries' line closing the matching annex table of SP-112_Draft_Report_v005.md, read 2026-08-04
  8. 365 people checked in and so kept their voting rights the '365 Entries' line closing the matching annex table of SP-112_Draft_Report_v005.md, read 2026-08-04
  9. 110 people registered to attend online the '110 Entries' line closing the matching annex table of SP-112_Draft_Report_v005.md, read 2026-08-04
  10. 120 documents of type 'CR pack' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and tdoctype='CR pack';", read 2026-08-04
  11. 53 documents of type 'WID new' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and tdoctype='WID new';", read 2026-08-04
  12. 29 documents of type 'report' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and tdoctype='report';", read 2026-08-04
  13. 28 documents of type 'draft TR' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and tdoctype='draft TR';", read 2026-08-04
  14. 20 documents of type 'discussion' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and tdoctype='discussion';", read 2026-08-04
  15. 18 documents of type 'LS in' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and tdoctype='LS in';", read 2026-08-04
  16. 14 documents of type 'CR' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and tdoctype='CR';", read 2026-08-04
  17. 13 documents of type 'other' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and tdoctype='other';", read 2026-08-04
  18. 10 documents of type 'SID new' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and tdoctype='SID new';", read 2026-08-04
  19. 9 documents of type 'SID revised' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and tdoctype='SID revised';", read 2026-08-04
  20. 199 documents ended the meeting 'approved' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and status='approved';", read 2026-08-04
  21. 68 documents ended the meeting 'noted' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and status='noted';", read 2026-08-04
  22. 60 documents ended the meeting 'revised' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and status='revised';", read 2026-08-04
  23. 5 documents ended the meeting 'postponed' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and status='postponed';", read 2026-08-04
  24. 4 documents ended the meeting 'endorsed' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and status='endorsed';", read 2026-08-04
  25. 4 documents ended the meeting 'partially approved' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and status='partially approved';", read 2026-08-04
  26. 2 documents ended the meeting 'merged' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and status='merged';", read 2026-08-04
  27. 2 documents ended the meeting 'withdrawn' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and status='withdrawn';", read 2026-08-04
  28. 1 document ended the meeting 'agreed' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and status='agreed';", read 2026-08-04
  29. 1 document ended the meeting 'replied to' sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and status='replied to';", read 2026-08-04
  30. 162 documents carry no release sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and (release is null or release='');", read 2026-08-04
  31. 140 documents belong to Rel-20 sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and release='Rel-20';", read 2026-08-04
  32. 26 documents belong to Rel-19 sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and release='Rel-19';", read 2026-08-04
  33. 11 documents belong to Rel-21 sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and release='Rel-21';", read 2026-08-04
  34. 5 documents belong to Rel-18 sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and release='Rel-18';", read 2026-08-04
  35. 2 documents belong to Rel-17 sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and release='Rel-17';", read 2026-08-04
  36. 66 documents came from SA WG5 sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and source="SA WG5";", read 2026-08-04
  37. 57 documents came from SA WG3 sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and source="SA WG3";", read 2026-08-04
  38. 41 documents came from SA WG6 sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and source="SA WG6";", read 2026-08-04
  39. 33 documents came from SA WG2 sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and source="SA WG2";", read 2026-08-04
  40. 28 documents came from SA WG4 sqlite3 /var/www/whatthespec.net/data/database/api/api.sqlite "select count(*) from tdoc where meeting='SP-112' and source="SA WG4";", read 2026-08-04
  41. 'It was decided to ask SA WG4, SA WG5 and SA WG6 to also submit ToRs to the next TSG SA meeting and SA WG1, SA WG2 and SA WG3 may update or re-submit the same ToR proposals.' found character for character in SP-112_Draft_Report_v005.md, which is a draft and says so on its cover, read 2026-08-04
  42. The report's own document list holds 346 entries the '346 Entries' line closing the matching annex table of SP-112_Draft_Report_v005.md, read 2026-08-04
  43. 487 change requests in packs belong to Rel-20 the crrelease field of every entry of the crsinpack list inside the line column of tdoc for meeting='SP-112', read 2026-08-04
  44. 143 change requests in packs belong to Rel-19 the crrelease field of every entry of the crsinpack list inside the line column of tdoc for meeting='SP-112', read 2026-08-04
  45. 36 change requests in packs belong to Rel-18 the crrelease field of every entry of the crsinpack list inside the line column of tdoc for meeting='SP-112', read 2026-08-04
  46. 7 change requests in packs belong to Rel-17 the crrelease field of every entry of the crsinpack list inside the line column of tdoc for meeting='SP-112', read 2026-08-04
  47. SP-112 draft meeting report, version 0.0.5
  48. 26 specifications and reports were approved the '26 Entries' line closing the matching annex table of SP-112_Draft_Report_v005.md, read 2026-08-04
  49. 6 specifications and reports were noted for information the '6 Entries' line closing the matching annex table of SP-112_Draft_Report_v005.md, read 2026-08-04
  50. 30 new work items and study items were approved the '30 Entries' line closing the matching annex table of SP-112_Draft_Report_v005.md, read 2026-08-04
  51. 11 work items and study items were updated the '11 Entries' line closing the matching annex table of SP-112_Draft_Report_v005.md, read 2026-08-04
  52. 18 incoming liaison statements the '18 Entries' line closing the matching annex table of SP-112_Draft_Report_v005.md, read 2026-08-04
  53. 2 outgoing liaison statements the '2 Entries' line closing the matching annex table of SP-112_Draft_Report_v005.md, read 2026-08-04
  54. No work item exception sheets were approved annex F.2 of SP-112_Draft_Report_v005.md, which reads 'No entries found.', read 2026-08-04
  55. 284 of the 346 documents get a paragraph of their own in the report distinct document numbers matching the pattern 'SP-nnnnnn (' at the start of a line in SP-112_Draft_Report_v005.md. The rest appear only in the annex tables, read 2026-08-04
  56. 40 times the report says 'There were no contributions to this agenda item.' occurrences of the sentence 'There were no contributions to this agenda item.' in SP-112_Draft_Report_v005.md. It counts sub-items as well as top-level items, read 2026-08-04
  57. 224 of the meeting's documents have a machine-written summary rows of table tdoc_summaries in summaries.db where meeting_id='TSGS_112_Singapore_2026-06', read 2026-08-04

Every source this course is built on

Check yourself

Answers appear when you pick one, with where they come from.

Q2.1 How many documents were handled at SA#112?

Q2.2 The ten status words account for how many of the meeting's documents?

Q2.3 The change requests of this meeting can be counted two ways, and the two totals differ. Why are both right?

Q2.4 What is the largest single kind of document at this meeting?

Q2.5 How many of the meeting's documents carry no release at all?

This chapter was built from a source register generated 2026-08-04. A fresher build of the register may hold different numbers.