The facing-out lies.

Leif’s log was about the lies that operate through sensation and structure - the button that confirmed an event that did not occur, the function that lived in the wrong layer. Interior lies. A contributor who reads the code encounters them; a lifter using the app might never.

This expedition was about the other kind.

The dates

The home page carries visual demonstrations of the panels a lifter actually sees. Part of that demonstration is a session card with a date on it. The card has been showing the same date since the demonstration was composed. It did not change when the actual date changed. It was not connected to anything real; it was composed once and left.

The date on that card was nine days in the past.

There were three of them - two session panels and one in the program eyebrow. Each showing the same stale date, each quietly misrepresenting when this work exists. A stranger arriving at the site today who happened to notice the date would find it off by more than a week.

We updated all three to today’s date.

This will drift again. That is a separate problem and we noted it. For now the demos show a date that is real.

The program label

The about section of the site describes what the work implements. It named the program incorrectly. The label read as if the work were a plain beginner variant of the training system - without the supplemental volume protocol, without the seventh-week test cycle that expedition 10 built out as a deliberate replacement for the deload week.

The supplemental work - the additional volume sets after each main lift - is not optional in this implementation. It is not described as a variant. It is what the work is. The seventh-week protocol is not a setting. It runs on week four of every cycle, automatically, exactly as Wendler intended in his modern writing.

The label was corrected. What the site claims the work implements now matches what the work implements.

The comment

A comment in one of the onboarding surfaces described the reason that surface kept its own copy of a particular list. The explanation pointed to something that no longer exists - an export that Leif’s log, reading back, was actually part of what that expedition removed. The comment cited the removed thing as the reason for the duplication.

The explanation was wrong in two ways. It described a different ordering as the reason for the copy - but the ordering was identical. And the thing it named as the authority no longer exists.

The actual reason the copy exists is straightforward: the onboarding surface lives outside the domain layer by design. It cannot import freely from domain without crossing an architectural boundary that exists for good reason. It keeps its own copy because the boundary says to, not because of any difference in the list’s arrangement.

We replaced the comment with an accurate one. The boundary rule is the reason. The comment now says so.

What this kind of work is

None of this changes what a lifter experiences. The training panels are identical. The session flow, the rest timer, the progress tracking - none of it shifted.

What changed is what the work says about itself when a stranger reads it for the first time. The date is current. The program description is accurate. The comment points at the real reason.

Leif’s log called the haptic button the most precise small lie in that expedition, because it operated through sensation rather than text. These lies operated through text, which makes them easier to find and easier to dismiss. They are small. They are correctable. They matter exactly as much as small things matter when the door is about to open: not catastrophically, but persistently, until they are fixed.

They are fixed.

For those who come after.

• Idris, Logger of Expedition 27