Logbook Sync, Replay Coverage, and the Next Device Boards

The last update was mostly about making replay and cross-device reliability more credible.

This next step is about what that foundation is now enabling.

The biggest practical change is that Apex Infinity can now move real logged activities from Core into Companion in a way that is finally useful enough to build on. At the same time, replay coverage has widened enough that the import path can be exercised against more than one activity profile, and the current Edge hardware investigation has reached the point where the right move is to change boards instead of spending unlimited time on a fragile display stack.

That is a very healthy kind of progress. It means more of the system is moving from bench-only validation into product behavior.

Core-to-Companion logbook sync is now real

One of the most important outcomes in this round is that logbook import from Core into Companion is now working end to end.

That does not mean the logbook story is finished. It does mean the basic path is real:

• Core can hold completed activity logs
• Companion can discover available slots
• Companion can import the sessions
• and the imported data can now be exported again as full raw CSV instead of being trapped inside an incomplete intermediate view

That matters because the logbook is one of the places where Apex Infinity stops feeling like a collection of live device surfaces and starts feeling like a system with memory.

Golden replay coverage is getting much stronger

This same round also pushed replay coverage forward in a more useful way.

Instead of validating against one replay file and assuming the rest of the import and presentation path would behave similarly, the system is now being exercised against a broader set of golden activity samples. That gives the platform much better coverage for:

• higher-profile scenarios
• lower-profile scenarios
• mid-profile sessions
• and shorter activity profiles

That kind of variety matters because activity logging and summary behavior should not only look correct for one idealized descent. The platform needs to stay coherent across different activity shapes if the logbook is going to become trustworthy.

Companion logbook behavior is much closer to product quality

On the Companion side, the logbook is no longer only importing data. It is starting to behave more like a real management surface.

This round added:

• cleaner imported session labeling based on the actual session number
• per-session and multi-select CSV export
• local delete behavior
• improved active use-oriented summary presentation
• and sync behavior that skips sessions already imported locally instead of pulling everything again every time

That is still not the final logbook product. There is still work to do around date fidelity, lifecycle management, and broader import support. But the difference between “offload data exists” and “the user can actually use the imported logbook” is significant, and the project has crossed that line.

The next data-quality work is now obvious

As soon as the import path started behaving, the next set of problems became much easier to see.

That is exactly what should happen in a healthy system. Once transport stops being the blocker, data quality becomes the real question.

The current next steps are straightforward:

• preserve trustworthy activity dates from Core instead of falling back to the import date
• continue tightening summary calculations where replay data exposes odd spikes or incomplete fields
• and expand the logbook to support imports from third-party devices such as FlySight and Dekunu

Those are much better problems to have than “the sync path does not work.”

Edge hardware work reached a useful decision point

Another important outcome in this round is that the current rectangular Edge hardware path is no longer being treated as an open-ended firmware science project.

The display investigation produced useful evidence:

• low-level panel writes can be stable
• some simpler LVGL paths can be stable
• but the current Hosyond-based live update path is still fragile enough that it is not the right place to spend indefinite time

That led to a practical hardware decision.

The next Edge bring-up will start with a new Waveshare device processor-cloud storage-Touch-LCD-2.8B, which is much closer to the Classic device stack and should reduce the amount of display-driver variability the platform has to carry. A Waveshare display hardware rectangular board is also queued behind it for evaluation, but the safer and more conservative next move is the 2.8B path first.

That is the kind of tradeoff that keeps development honest. Sometimes the correct technical answer is not “debug the current board forever.” Sometimes it is “move to hardware with a better chance of becoming a reliable product.”

Why this phase matters

This is not a flashy feature-only update.

It is the kind of platform work that makes later user-facing features credible:

• replay is broad enough to test import behavior across different activity shapes
• logbook import is real enough to exercise actual user flows
• Companion can now export the imported data in a useful form
• and the Wrist hardware path is being steered toward a more defensible foundation

That combination matters because Apex Infinity is trying to become more than a live device display. It is trying to become a connected system that can record, restore, validate, and explain the session experience across multiple surfaces.

What is next

The next steps are clear from here.

Core still needs to provide a trustworthy session timestamp so imported logbook entries carry the real session date. Companion needs continued refinement around lifecycle, metrics, and additional import sources. And the next generation of Edge hardware needs to be brought up on the new Waveshare boards so the rectangular Wrist path can move forward on a stronger base.

That is not the end of the logbook story or the end of the device story.

But it is a meaningful shift: Apex Infinity now has a real import path, stronger replay coverage behind it, and a clearer plan for the hardware that still needs to catch up.