Large-family-tree field guide
Keep the whole record.
Show only what helps.
A family tree with thousands of people becomes useful when relatives can begin with someone familiar, reveal the relevant branch and inspect the path behind an answer. Size is a navigation and stewardship problem—not a reason to publish the family record.
10,000 people tested without using a real family record.
Kinonicle’s automated rehearsal validates a broad synthetic GEDCOM and separately traverses a deliberately extreme 5,000-generation lineage. The test is a regression guardrail, not a device-speed claim.
A large tree is not one enormous poster
The complete record and the useful view are different things. The archive may contain thousands of people, while a relative’s immediate question concerns one person, a few generations or a path between two branches.
Trying to display everything at once makes names unreadable and relationships difficult to follow. A better interface keeps the complete private structure available, then creates a smaller working view around the question at hand.
Use a six-part navigation method
- 1
Search before scanning
Find a known person by name, then make that person the centre of the current view. A large tree should not require someone to visually scan thousands of nodes.
- 2
Reveal a useful window
Show the generations and branches needed for the present question. Keep the rest available without placing the entire archive on screen at once.
- 3
Follow one recorded path
When two people ask how they are related, show the person-to-person path and let the family inspect each parent, child or spouse link beneath the label.
- 4
Check structure before appearance
After an import, compare counts and inspect familiar parent-child, spouse and sibling groups. A tidy diagram cannot prove that the underlying links are correct.
- 5
Resolve duplicates deliberately
Treat similar names, dates or places as clues rather than proof. Keep unresolved people separate until a steward can review sources and family context.
- 6
Preserve language and context
Store the genealogical path as the structural record while retaining the names, scripts and kinship terms the family actually uses.
Check a large import with known paths
Counts are the first checkpoint, not the last. Record the source application, export date, file checksum, person count and relationship count. After import, compare the report and inspect several paths the steward already knows.
- A direct ancestral line over at least three generations.
- A household with spouses, siblings and children.
- A relationship path connecting two distant but familiar branches.
- A person whose name uses another script, alias or generational convention.
A mismatch should remain visible until it is explained. Do not make the numbers agree by silently merging people or inventing a missing link.
Scale stewardship with the record
A bigger archive creates more opportunities for conflicting memories, accidental duplicates and sensitive living-person information. Give each relative an individual account, let contributions remain reviewable and make the source or reasoning behind a material change understandable.
Define who resolves routine corrections, how disputed facts are represented, when exports are created and who can continue the work if the primary steward steps down. More contributors should improve the evidence without turning the trusted record into an unrestricted public tree.
What Kinonicle rehearses
The automated gate passed on 3 August 2026. It uses generated names and relationships only; it reads no production family, customer or private archive data.
What remains to be proven
Import capacity is not complete experience evidence
Before a broad public launch, Kinonicle must still render a synthetic multi-thousand-person archive in the production build and record first interaction, search and focus, pan and zoom, branch controls, relationship fitting, browser memory and long tasks on a representative phone and desktop.
This page intentionally makes no claim about a particular device speed, unlimited family size or flawless import of every GEDCOM field. The current public evidence is the reproducible synthetic parser and traversal rehearsal described above.