Write the comparison table in words people actually use #73

Closed
paulhitt wants to merge 3 commits from 12-plain-comparison into main
paulhitt commented 2026-08-31 21:12:43 +00:00 (Migrated from gitlab.com)

Closes #11
Closes #12

Plain-language pass over every row heading and every line of small print on /comparison. Protocol names move to the sources list at the foot of the page, where a reader who wants to go and check something needs the real name; nothing in the table itself.

Two corrections ride along:

  • Signal was scored No on the ongoing-ratchet row and that is wrong — they shipped a sparse post-quantum ratchet in 2025. Now Yes, and "Reading the table fairly" no longer claims Apple is ahead of everyone.
  • "You can install it today" still said the iPhone app was in build and the Android engine had started. Android 0.7 has been downloadable since this morning.

Depends on flamenet-e2e!5. spec.leaf is regenerated from §15 in the engine repo, so check-doc-pages.py here goes red until that one is on main.

Closes #11 Closes #12 Plain-language pass over every row heading and every line of small print on `/comparison`. Protocol names move to the sources list at the foot of the page, where a reader who wants to go and check something needs the real name; nothing in the table itself. Two corrections ride along: - **Signal was scored No on the ongoing-ratchet row and that is wrong** — they shipped a sparse post-quantum ratchet in 2025. Now Yes, and "Reading the table fairly" no longer claims Apple is ahead of everyone. - **"You can install it today"** still said the iPhone app was in build and the Android engine had started. Android 0.7 has been downloadable since this morning. Depends on flamenet-e2e!5. `spec.leaf` is regenerated from §15 in the engine repo, so `check-doc-pages.py` here goes red until that one is on `main`.
paulhitt commented 2026-08-31 21:14:24 +00:00 (Migrated from gitlab.com)

added 1 commit

  • 81623b57 - Do not say Apple's protocol was proved correct

Compare with previous version

added 1 commit <ul><li>81623b57 - Do not say Apple&#39;s protocol was proved correct</li></ul> [Compare with previous version](/paulhitt/flamenet-relay/-/merge_requests/10/diffs?diff_id=1995295540&start_sha=71231003d79e502a245386e8437048bc073f5c68)
paulhitt commented 2026-08-31 21:15:05 +00:00 (Migrated from gitlab.com)

added 1 commit

  • db387345 - The lede counted three and the table has five

Compare with previous version

added 1 commit <ul><li>db387345 - The lede counted three and the table has five</li></ul> [Compare with previous version](/paulhitt/flamenet-relay/-/merge_requests/10/diffs?diff_id=1995296941&start_sha=81623b575d21e55f50932d003827559aab00b90b)
paulhitt commented 2026-09-02 23:52:29 +00:00 (Migrated from gitlab.com)

Held back, not forgotten. The rebase conflicts in comparison.leaf, which is hand-written rather than generated, and the two sides disagree on more than wording: main carries newer facts (the post-quantum ratchet row is now "Android now", the audit row is "Planned"), while this branch rewrites the phrasing and gives that row a different number of cells than the table on main has. Merging it mechanically would either revert the facts or break the column alignment.

It is public copy and affects nothing anyone is testing, so it is the wrong thing to resolve at speed. Worth ten minutes with the rendered page open.

Held back, not forgotten. The rebase conflicts in `comparison.leaf`, which is hand-written rather than generated, and the two sides disagree on more than wording: main carries newer facts (the post-quantum ratchet row is now "Android now", the audit row is "Planned"), while this branch rewrites the phrasing and gives that row a different number of cells than the table on main has. Merging it mechanically would either revert the facts or break the column alignment. It is public copy and affects nothing anyone is testing, so it is the wrong thing to resolve at speed. Worth ten minutes with the rendered page open.
paulhitt commented 2026-09-05 20:17:42 +00:00 (Migrated from gitlab.com)

mentioned in merge request !68

mentioned in merge request !68
paulhitt commented 2026-09-07 11:58:35 +00:00 (Migrated from gitlab.com)

Closing as superseded rather than merging.

Two branches rewrote comparison.leaf in plain words, and the other one landed (main, 1b915b0). Main's version carries the same intent — plain-word headings, and a The real names, if you want to go and check section that keeps the protocol names available without leading with them — so nothing here is lost by closing it.

The spec.leaf half is a generated file. This branch carries source-sha256: 386d1af664af3a4a; main is already at 3a9247e4b19fd281, generated from a newer .e2e-ref (d2ebcdf) that this branch predates entirely. Merging would move a generated page backwards.

Related: this is the same shape as the collision graph note — GitLab reported this MR mergeable throughout, because "mergeable" compares a branch to main and never to the other MR editing the same file.

Closing as superseded rather than merging. Two branches rewrote `comparison.leaf` in plain words, and the other one landed (main, `1b915b0`). Main's version carries the same intent — plain-word headings, and a *The real names, if you want to go and check* section that keeps the protocol names available without leading with them — so nothing here is lost by closing it. The `spec.leaf` half is a **generated** file. This branch carries `source-sha256: 386d1af664af3a4a`; main is already at `3a9247e4b19fd281`, generated from a newer `.e2e-ref` (`d2ebcdf`) that this branch predates entirely. Merging would move a generated page *backwards*. Related: this is the same shape as the collision graph note — GitLab reported this MR mergeable throughout, because "mergeable" compares a branch to main and never to the other MR editing the same file.
paulhitt (Migrated from gitlab.com) closed this pull request 2026-09-07 11:58:35 +00:00
paulhitt commented 2026-09-09 18:59:32 +00:00 (Migrated from gitlab.com)

changed the description

changed the description

Pull request closed

This pull request cannot be reopened because the branch was deleted.
Sign in to join this conversation.